WorldWideScience

Sample records for unsupervised hierarchical clustering

  1. Unsupervised active learning based on hierarchical graph-theoretic clustering.

    Science.gov (United States)

    Hu, Weiming; Hu, Wei; Xie, Nianhua; Maybank, Steve

    2009-10-01

    Most existing active learning approaches are supervised. Supervised active learning has the following problems: inefficiency in dealing with the semantic gap between the distribution of samples in the feature space and their labels, lack of ability in selecting new samples that belong to new categories that have not yet appeared in the training samples, and lack of adaptability to changes in the semantic interpretation of sample categories. To tackle these problems, we propose an unsupervised active learning framework based on hierarchical graph-theoretic clustering. In the framework, two promising graph-theoretic clustering algorithms, namely, dominant-set clustering and spectral clustering, are combined in a hierarchical fashion. Our framework has some advantages, such as ease of implementation, flexibility in architecture, and adaptability to changes in the labeling. Evaluations on data sets for network intrusion detection, image classification, and video classification have demonstrated that our active learning framework can effectively reduce the workload of manual classification while maintaining a high accuracy of automatic classification. It is shown that, overall, our framework outperforms the support-vector-machine-based supervised active learning, particularly in terms of dealing much more efficiently with new samples whose categories have not yet appeared in the training samples.

  2. Hierarchical Adaptive Means (HAM) clustering for hardware-efficient, unsupervised and real-time spike sorting.

    Science.gov (United States)

    Paraskevopoulou, Sivylla E; Wu, Di; Eftekhar, Amir; Constandinou, Timothy G

    2014-09-30

    This work presents a novel unsupervised algorithm for real-time adaptive clustering of neural spike data (spike sorting). The proposed Hierarchical Adaptive Means (HAM) clustering method combines centroid-based clustering with hierarchical cluster connectivity to classify incoming spikes using groups of clusters. It is described how the proposed method can adaptively track the incoming spike data without requiring any past history, iteration or training and autonomously determines the number of spike classes. Its performance (classification accuracy) has been tested using multiple datasets (both simulated and recorded) achieving a near-identical accuracy compared to k-means (using 10-iterations and provided with the number of spike classes). Also, its robustness in applying to different feature extraction methods has been demonstrated by achieving classification accuracies above 80% across multiple datasets. Last but crucially, its low complexity, that has been quantified through both memory and computation requirements makes this method hugely attractive for future hardware implementation. Copyright © 2014 Elsevier B.V. All rights reserved.

  3. Statistical Significance for Hierarchical Clustering

    Science.gov (United States)

    Kimes, Patrick K.; Liu, Yufeng; Hayes, D. Neil; Marron, J. S.

    2017-01-01

    Summary Cluster analysis has proved to be an invaluable tool for the exploratory and unsupervised analysis of high dimensional datasets. Among methods for clustering, hierarchical approaches have enjoyed substantial popularity in genomics and other fields for their ability to simultaneously uncover multiple layers of clustering structure. A critical and challenging question in cluster analysis is whether the identified clusters represent important underlying structure or are artifacts of natural sampling variation. Few approaches have been proposed for addressing this problem in the context of hierarchical clustering, for which the problem is further complicated by the natural tree structure of the partition, and the multiplicity of tests required to parse the layers of nested clusters. In this paper, we propose a Monte Carlo based approach for testing statistical significance in hierarchical clustering which addresses these issues. The approach is implemented as a sequential testing procedure guaranteeing control of the family-wise error rate. Theoretical justification is provided for our approach, and its power to detect true clustering structure is illustrated through several simulation studies and applications to two cancer gene expression datasets. PMID:28099990

  4. Massively parallel unsupervised single-particle cryo-EM data clustering via statistical manifold learning.

    Directory of Open Access Journals (Sweden)

    Jiayi Wu

    Full Text Available Structural heterogeneity in single-particle cryo-electron microscopy (cryo-EM data represents a major challenge for high-resolution structure determination. Unsupervised classification may serve as the first step in the assessment of structural heterogeneity. However, traditional algorithms for unsupervised classification, such as K-means clustering and maximum likelihood optimization, may classify images into wrong classes with decreasing signal-to-noise-ratio (SNR in the image data, yet demand increased computational costs. Overcoming these limitations requires further development of clustering algorithms for high-performance cryo-EM data processing. Here we introduce an unsupervised single-particle clustering algorithm derived from a statistical manifold learning framework called generative topographic mapping (GTM. We show that unsupervised GTM clustering improves classification accuracy by about 40% in the absence of input references for data with lower SNRs. Applications to several experimental datasets suggest that our algorithm can detect subtle structural differences among classes via a hierarchical clustering strategy. After code optimization over a high-performance computing (HPC environment, our software implementation was able to generate thousands of reference-free class averages within hours in a massively parallel fashion, which allows a significant improvement on ab initio 3D reconstruction and assists in the computational purification of homogeneous datasets for high-resolution visualization.

  5. Massively parallel unsupervised single-particle cryo-EM data clustering via statistical manifold learning.

    Science.gov (United States)

    Wu, Jiayi; Ma, Yong-Bei; Congdon, Charles; Brett, Bevin; Chen, Shuobing; Xu, Yaofang; Ouyang, Qi; Mao, Youdong

    2017-01-01

    Structural heterogeneity in single-particle cryo-electron microscopy (cryo-EM) data represents a major challenge for high-resolution structure determination. Unsupervised classification may serve as the first step in the assessment of structural heterogeneity. However, traditional algorithms for unsupervised classification, such as K-means clustering and maximum likelihood optimization, may classify images into wrong classes with decreasing signal-to-noise-ratio (SNR) in the image data, yet demand increased computational costs. Overcoming these limitations requires further development of clustering algorithms for high-performance cryo-EM data processing. Here we introduce an unsupervised single-particle clustering algorithm derived from a statistical manifold learning framework called generative topographic mapping (GTM). We show that unsupervised GTM clustering improves classification accuracy by about 40% in the absence of input references for data with lower SNRs. Applications to several experimental datasets suggest that our algorithm can detect subtle structural differences among classes via a hierarchical clustering strategy. After code optimization over a high-performance computing (HPC) environment, our software implementation was able to generate thousands of reference-free class averages within hours in a massively parallel fashion, which allows a significant improvement on ab initio 3D reconstruction and assists in the computational purification of homogeneous datasets for high-resolution visualization.

  6. Inferring hierarchical clustering structures by deterministic annealing

    International Nuclear Information System (INIS)

    Hofmann, T.; Buhmann, J.M.

    1996-01-01

    The unsupervised detection of hierarchical structures is a major topic in unsupervised learning and one of the key questions in data analysis and representation. We propose a novel algorithm for the problem of learning decision trees for data clustering and related problems. In contrast to many other methods based on successive tree growing and pruning, we propose an objective function for tree evaluation and we derive a non-greedy technique for tree growing. Applying the principles of maximum entropy and minimum cross entropy, a deterministic annealing algorithm is derived in a meanfield approximation. This technique allows us to canonically superimpose tree structures and to fit parameters to averaged or open-quote fuzzified close-quote trees

  7. Hierarchical Bayesian nonparametric mixture models for clustering with variable relevance determination.

    Science.gov (United States)

    Yau, Christopher; Holmes, Chris

    2011-07-01

    We propose a hierarchical Bayesian nonparametric mixture model for clustering when some of the covariates are assumed to be of varying relevance to the clustering problem. This can be thought of as an issue in variable selection for unsupervised learning. We demonstrate that by defining a hierarchical population based nonparametric prior on the cluster locations scaled by the inverse covariance matrices of the likelihood we arrive at a 'sparsity prior' representation which admits a conditionally conjugate prior. This allows us to perform full Gibbs sampling to obtain posterior distributions over parameters of interest including an explicit measure of each covariate's relevance and a distribution over the number of potential clusters present in the data. This also allows for individual cluster specific variable selection. We demonstrate improved inference on a number of canonical problems.

  8. Hierarchical Aligned Cluster Analysis for Temporal Clustering of Human Motion.

    Science.gov (United States)

    Zhou, Feng; De la Torre, Fernando; Hodgins, Jessica K

    2013-03-01

    Temporal segmentation of human motion into plausible motion primitives is central to understanding and building computational models of human motion. Several issues contribute to the challenge of discovering motion primitives: the exponential nature of all possible movement combinations, the variability in the temporal scale of human actions, and the complexity of representing articulated motion. We pose the problem of learning motion primitives as one of temporal clustering, and derive an unsupervised hierarchical bottom-up framework called hierarchical aligned cluster analysis (HACA). HACA finds a partition of a given multidimensional time series into m disjoint segments such that each segment belongs to one of k clusters. HACA combines kernel k-means with the generalized dynamic time alignment kernel to cluster time series data. Moreover, it provides a natural framework to find a low-dimensional embedding for time series. HACA is efficiently optimized with a coordinate descent strategy and dynamic programming. Experimental results on motion capture and video data demonstrate the effectiveness of HACA for segmenting complex motions and as a visualization tool. We also compare the performance of HACA to state-of-the-art algorithms for temporal clustering on data of a honey bee dance. The HACA code is available online.

  9. Similarity maps and hierarchical clustering for annotating FT-IR spectral images.

    Science.gov (United States)

    Zhong, Qiaoyong; Yang, Chen; Großerüschkamp, Frederik; Kallenbach-Thieltges, Angela; Serocka, Peter; Gerwert, Klaus; Mosig, Axel

    2013-11-20

    Unsupervised segmentation of multi-spectral images plays an important role in annotating infrared microscopic images and is an essential step in label-free spectral histopathology. In this context, diverse clustering approaches have been utilized and evaluated in order to achieve segmentations of Fourier Transform Infrared (FT-IR) microscopic images that agree with histopathological characterization. We introduce so-called interactive similarity maps as an alternative annotation strategy for annotating infrared microscopic images. We demonstrate that segmentations obtained from interactive similarity maps lead to similarly accurate segmentations as segmentations obtained from conventionally used hierarchical clustering approaches. In order to perform this comparison on quantitative grounds, we provide a scheme that allows to identify non-horizontal cuts in dendrograms. This yields a validation scheme for hierarchical clustering approaches commonly used in infrared microscopy. We demonstrate that interactive similarity maps may identify more accurate segmentations than hierarchical clustering based approaches, and thus are a viable and due to their interactive nature attractive alternative to hierarchical clustering. Our validation scheme furthermore shows that performance of hierarchical two-means is comparable to the traditionally used Ward's clustering. As the former is much more efficient in time and memory, our results suggest another less resource demanding alternative for annotating large spectral images.

  10. Hierarchical clustering of HPV genotype patterns in the ASCUS-LSIL triage study

    Science.gov (United States)

    Wentzensen, Nicolas; Wilson, Lauren E.; Wheeler, Cosette M.; Carreon, Joseph D.; Gravitt, Patti E.; Schiffman, Mark; Castle, Philip E.

    2010-01-01

    Anogenital cancers are associated with about 13 carcinogenic HPV types in a broader group that cause cervical intraepithelial neoplasia (CIN). Multiple concurrent cervical HPV infections are common which complicate the attribution of HPV types to different grades of CIN. Here we report the analysis of HPV genotype patterns in the ASCUS-LSIL triage study using unsupervised hierarchical clustering. Women who underwent colposcopy at baseline (n = 2780) were grouped into 20 disease categories based on histology and cytology. Disease groups and HPV genotypes were clustered using complete linkage. Risk of 2-year cumulative CIN3+, viral load, colposcopic impression, and age were compared between disease groups and major clusters. Hierarchical clustering yielded four major disease clusters: Cluster 1 included all CIN3 histology with abnormal cytology; Cluster 2 included CIN3 histology with normal cytology and combinations with either CIN2 or high-grade squamous intraepithelial lesion (HSIL) cytology; Cluster 3 included older women with normal or low grade histology/cytology and low viral load; Cluster 4 included younger women with low grade histology/cytology, multiple infections, and the highest viral load. Three major groups of HPV genotypes were identified: Group 1 included only HPV16; Group 2 included nine carcinogenic types plus non-carcinogenic HPV53 and HPV66; and Group 3 included non-carcinogenic types plus carcinogenic HPV33 and HPV45. Clustering results suggested that colposcopy missed a prevalent precancer in many women with no biopsy/normal histology and HSIL. This result was confirmed by an elevated 2-year risk of CIN3+ in these groups. Our novel approach to study multiple genotype infections in cervical disease using unsupervised hierarchical clustering can address complex genotype distributions on a population level. PMID:20959485

  11. Hierarchical Multiple Markov Chain Model for Unsupervised Texture Segmentation

    Czech Academy of Sciences Publication Activity Database

    Scarpa, G.; Gaetano, R.; Haindl, Michal; Zerubia, J.

    2009-01-01

    Roč. 18, č. 8 (2009), s. 1830-1843 ISSN 1057-7149 R&D Projects: GA ČR GA102/08/0593 EU Projects: European Commission(XE) 507752 - MUSCLE Institutional research plan: CEZ:AV0Z10750506 Keywords : Classification * texture analysis * segmentation * hierarchical image models * Markov process Subject RIV: BD - Theory of Information Impact factor: 2.848, year: 2009 http://library.utia.cas.cz/separaty/2009/RO/haindl-hierarchical multiple markov chain model for unsupervised texture segmentation.pdf

  12. Unsupervised Cryo-EM Data Clustering through Adaptively Constrained K-Means Algorithm.

    Science.gov (United States)

    Xu, Yaofang; Wu, Jiayi; Yin, Chang-Cheng; Mao, Youdong

    2016-01-01

    In single-particle cryo-electron microscopy (cryo-EM), K-means clustering algorithm is widely used in unsupervised 2D classification of projection images of biological macromolecules. 3D ab initio reconstruction requires accurate unsupervised classification in order to separate molecular projections of distinct orientations. Due to background noise in single-particle images and uncertainty of molecular orientations, traditional K-means clustering algorithm may classify images into wrong classes and produce classes with a large variation in membership. Overcoming these limitations requires further development on clustering algorithms for cryo-EM data analysis. We propose a novel unsupervised data clustering method building upon the traditional K-means algorithm. By introducing an adaptive constraint term in the objective function, our algorithm not only avoids a large variation in class sizes but also produces more accurate data clustering. Applications of this approach to both simulated and experimental cryo-EM data demonstrate that our algorithm is a significantly improved alterative to the traditional K-means algorithm in single-particle cryo-EM analysis.

  13. Application of cluster analysis and unsupervised learning to multivariate tissue characterization

    International Nuclear Information System (INIS)

    Momenan, R.; Insana, M.F.; Wagner, R.F.; Garra, B.S.; Loew, M.H.

    1987-01-01

    This paper describes a procedure for classifying tissue types from unlabeled acoustic measurements (data type unknown) using unsupervised cluster analysis. These techniques are being applied to unsupervised ultrasonic image segmentation and tissue characterization. The performance of a new clustering technique is measured and compared with supervised methods, such as a linear Bayes classifier. In these comparisons two objectives are sought: a) How well does the clustering method group the data?; b) Do the clusters correspond to known tissue classes? The first question is investigated by a measure of cluster similarity and dispersion. The second question involves a comparison with a supervised technique using labeled data

  14. Performance Analysis of Unsupervised Clustering Methods for Brain Tumor Segmentation

    Directory of Open Access Journals (Sweden)

    Tushar H Jaware

    2013-10-01

    Full Text Available Medical image processing is the most challenging and emerging field of neuroscience. The ultimate goal of medical image analysis in brain MRI is to extract important clinical features that would improve methods of diagnosis & treatment of disease. This paper focuses on methods to detect & extract brain tumour from brain MR images. MATLAB is used to design, software tool for locating brain tumor, based on unsupervised clustering methods. K-Means clustering algorithm is implemented & tested on data base of 30 images. Performance evolution of unsupervised clusteringmethods is presented.

  15. Neutrosophic Hierarchical Clustering Algoritms

    Directory of Open Access Journals (Sweden)

    Rıdvan Şahin

    2014-03-01

    Full Text Available Interval neutrosophic set (INS is a generalization of interval valued intuitionistic fuzzy set (IVIFS, whose the membership and non-membership values of elements consist of fuzzy range, while single valued neutrosophic set (SVNS is regarded as extension of intuitionistic fuzzy set (IFS. In this paper, we extend the hierarchical clustering techniques proposed for IFSs and IVIFSs to SVNSs and INSs respectively. Based on the traditional hierarchical clustering procedure, the single valued neutrosophic aggregation operator, and the basic distance measures between SVNSs, we define a single valued neutrosophic hierarchical clustering algorithm for clustering SVNSs. Then we extend the algorithm to classify an interval neutrosophic data. Finally, we present some numerical examples in order to show the effectiveness and availability of the developed clustering algorithms.

  16. Factored Translation with Unsupervised Word Clusters

    DEFF Research Database (Denmark)

    Rishøj, Christian; Søgaard, Anders

    2011-01-01

    Unsupervised word clustering algorithms — which form word clusters based on a measure of distributional similarity — have proven to be useful in providing beneficial features for various natural language processing tasks involving supervised learning. This work explores the utility of such word...... clusters as factors in statistical machine translation. Although some of the language pairs in this work clearly benefit from the factor augmentation, there is no consistent improvement in translation accuracy across the board. For all language pairs, the word clusters clearly improve translation for some...... proportion of the sentences in the test set, but has a weak or even detrimental effect on the rest. It is shown that if one could determine whether or not to use a factor when translating a given sentence, rather substantial improvements in precision could be achieved for all of the language pairs evaluated...

  17. Tune Your Brown Clustering, Please

    DEFF Research Database (Denmark)

    Derczynski, Leon; Chester, Sean; Bøgh, Kenneth Sejdenfaden

    2015-01-01

    Brown clustering, an unsupervised hierarchical clustering technique based on ngram mutual information, has proven useful in many NLP applications. However, most uses of Brown clustering employ the same default configuration; the appropriateness of this configuration has gone predominantly...

  18. Convex Clustering: An Attractive Alternative to Hierarchical Clustering

    Science.gov (United States)

    Chen, Gary K.; Chi, Eric C.; Ranola, John Michael O.; Lange, Kenneth

    2015-01-01

    The primary goal in cluster analysis is to discover natural groupings of objects. The field of cluster analysis is crowded with diverse methods that make special assumptions about data and address different scientific aims. Despite its shortcomings in accuracy, hierarchical clustering is the dominant clustering method in bioinformatics. Biologists find the trees constructed by hierarchical clustering visually appealing and in tune with their evolutionary perspective. Hierarchical clustering operates on multiple scales simultaneously. This is essential, for instance, in transcriptome data, where one may be interested in making qualitative inferences about how lower-order relationships like gene modules lead to higher-order relationships like pathways or biological processes. The recently developed method of convex clustering preserves the visual appeal of hierarchical clustering while ameliorating its propensity to make false inferences in the presence of outliers and noise. The solution paths generated by convex clustering reveal relationships between clusters that are hidden by static methods such as k-means clustering. The current paper derives and tests a novel proximal distance algorithm for minimizing the objective function of convex clustering. The algorithm separates parameters, accommodates missing data, and supports prior information on relationships. Our program CONVEXCLUSTER incorporating the algorithm is implemented on ATI and nVidia graphics processing units (GPUs) for maximal speed. Several biological examples illustrate the strengths of convex clustering and the ability of the proximal distance algorithm to handle high-dimensional problems. CONVEXCLUSTER can be freely downloaded from the UCLA Human Genetics web site at http://www.genetics.ucla.edu/software/ PMID:25965340

  19. Clustervision: Visual Supervision of Unsupervised Clustering.

    Science.gov (United States)

    Kwon, Bum Chul; Eysenbach, Ben; Verma, Janu; Ng, Kenney; De Filippi, Christopher; Stewart, Walter F; Perer, Adam

    2018-01-01

    Clustering, the process of grouping together similar items into distinct partitions, is a common type of unsupervised machine learning that can be useful for summarizing and aggregating complex multi-dimensional data. However, data can be clustered in many ways, and there exist a large body of algorithms designed to reveal different patterns. While having access to a wide variety of algorithms is helpful, in practice, it is quite difficult for data scientists to choose and parameterize algorithms to get the clustering results relevant for their dataset and analytical tasks. To alleviate this problem, we built Clustervision, a visual analytics tool that helps ensure data scientists find the right clustering among the large amount of techniques and parameters available. Our system clusters data using a variety of clustering techniques and parameters and then ranks clustering results utilizing five quality metrics. In addition, users can guide the system to produce more relevant results by providing task-relevant constraints on the data. Our visual user interface allows users to find high quality clustering results, explore the clusters using several coordinated visualization techniques, and select the cluster result that best suits their task. We demonstrate this novel approach using a case study with a team of researchers in the medical domain and showcase that our system empowers users to choose an effective representation of their complex data.

  20. Misty Mountain clustering: application to fast unsupervised flow cytometry gating

    Directory of Open Access Journals (Sweden)

    Sealfon Stuart C

    2010-10-01

    Full Text Available Abstract Background There are many important clustering questions in computational biology for which no satisfactory method exists. Automated clustering algorithms, when applied to large, multidimensional datasets, such as flow cytometry data, prove unsatisfactory in terms of speed, problems with local minima or cluster shape bias. Model-based approaches are restricted by the assumptions of the fitting functions. Furthermore, model based clustering requires serial clustering for all cluster numbers within a user defined interval. The final cluster number is then selected by various criteria. These supervised serial clustering methods are time consuming and frequently different criteria result in different optimal cluster numbers. Various unsupervised heuristic approaches that have been developed such as affinity propagation are too expensive to be applied to datasets on the order of 106 points that are often generated by high throughput experiments. Results To circumvent these limitations, we developed a new, unsupervised density contour clustering algorithm, called Misty Mountain, that is based on percolation theory and that efficiently analyzes large data sets. The approach can be envisioned as a progressive top-down removal of clouds covering a data histogram relief map to identify clusters by the appearance of statistically distinct peaks and ridges. This is a parallel clustering method that finds every cluster after analyzing only once the cross sections of the histogram. The overall run time for the composite steps of the algorithm increases linearly by the number of data points. The clustering of 106 data points in 2D data space takes place within about 15 seconds on a standard laptop PC. Comparison of the performance of this algorithm with other state of the art automated flow cytometry gating methods indicate that Misty Mountain provides substantial improvements in both run time and in the accuracy of cluster assignment. Conclusions

  1. Unsupervised clustering with spiking neurons by sparse temporal coding and multi-layer RBF networks

    NARCIS (Netherlands)

    S.M. Bohte (Sander); J.A. La Poutré (Han); J.N. Kok (Joost)

    2000-01-01

    textabstractWe demonstrate that spiking neural networks encoding information in spike times are capable of computing and learning clusters from realistic data. We show how a spiking neural network based on spike-time coding and Hebbian learning can successfully perform unsupervised clustering on

  2. Artificial immune kernel clustering network for unsupervised image segmentation

    Institute of Scientific and Technical Information of China (English)

    Wenlong Huang; Licheng Jiao

    2008-01-01

    An immune kernel clustering network (IKCN) is proposed based on the combination of the artificial immune network and the support vector domain description (SVDD) for the unsupervised image segmentation. In the network, a new antibody neighborhood and an adaptive learning coefficient, which is inspired by the long-term memory in cerebral cortices are presented. Starting from IKCN algorithm, we divide the image feature sets into subsets by the antibodies, and then map each subset into a high dimensional feature space by a mercer kernel, where each antibody neighborhood is represented as a support vector hypersphere. The clustering results of the local support vector hyperspheres are combined to yield a global clustering solution by the minimal spanning tree (MST), where a predefined number of clustering is not needed. We compare the proposed methods with two common clustering algorithms for the artificial synthetic data set and several image data sets, including the synthetic texture images and the SAR images, and encouraging experimental results are obtained.

  3. A novel unsupervised spike sorting algorithm for intracranial EEG.

    Science.gov (United States)

    Yadav, R; Shah, A K; Loeb, J A; Swamy, M N S; Agarwal, R

    2011-01-01

    This paper presents a novel, unsupervised spike classification algorithm for intracranial EEG. The method combines template matching and principal component analysis (PCA) for building a dynamic patient-specific codebook without a priori knowledge of the spike waveforms. The problem of misclassification due to overlapping classes is resolved by identifying similar classes in the codebook using hierarchical clustering. Cluster quality is visually assessed by projecting inter- and intra-clusters onto a 3D plot. Intracranial EEG from 5 patients was utilized to optimize the algorithm. The resulting codebook retains 82.1% of the detected spikes in non-overlapping and disjoint clusters. Initial results suggest a definite role of this method for both rapid review and quantitation of interictal spikes that could enhance both clinical treatment and research studies on epileptic patients.

  4. Unsupervised deep learning reveals prognostically relevant subtypes of glioblastoma.

    Science.gov (United States)

    Young, Jonathan D; Cai, Chunhui; Lu, Xinghua

    2017-10-03

    One approach to improving the personalized treatment of cancer is to understand the cellular signaling transduction pathways that cause cancer at the level of the individual patient. In this study, we used unsupervised deep learning to learn the hierarchical structure within cancer gene expression data. Deep learning is a group of machine learning algorithms that use multiple layers of hidden units to capture hierarchically related, alternative representations of the input data. We hypothesize that this hierarchical structure learned by deep learning will be related to the cellular signaling system. Robust deep learning model selection identified a network architecture that is biologically plausible. Our model selection results indicated that the 1st hidden layer of our deep learning model should contain about 1300 hidden units to most effectively capture the covariance structure of the input data. This agrees with the estimated number of human transcription factors, which is approximately 1400. This result lends support to our hypothesis that the 1st hidden layer of a deep learning model trained on gene expression data may represent signals related to transcription factor activation. Using the 3rd hidden layer representation of each tumor as learned by our unsupervised deep learning model, we performed consensus clustering on all tumor samples-leading to the discovery of clusters of glioblastoma multiforme with differential survival. One of these clusters contained all of the glioblastoma samples with G-CIMP, a known methylation phenotype driven by the IDH1 mutation and associated with favorable prognosis, suggesting that the hidden units in the 3rd hidden layer representations captured a methylation signal without explicitly using methylation data as input. We also found differentially expressed genes and well-known mutations (NF1, IDH1, EGFR) that were uniquely correlated with each of these clusters. Exploring these unique genes and mutations will allow us to

  5. UNSUPERVISED TRANSIENT LIGHT CURVE ANALYSIS VIA HIERARCHICAL BAYESIAN INFERENCE

    International Nuclear Information System (INIS)

    Sanders, N. E.; Soderberg, A. M.; Betancourt, M.

    2015-01-01

    Historically, light curve studies of supernovae (SNe) and other transient classes have focused on individual objects with copious and high signal-to-noise observations. In the nascent era of wide field transient searches, objects with detailed observations are decreasing as a fraction of the overall known SN population, and this strategy sacrifices the majority of the information contained in the data about the underlying population of transients. A population level modeling approach, simultaneously fitting all available observations of objects in a transient sub-class of interest, fully mines the data to infer the properties of the population and avoids certain systematic biases. We present a novel hierarchical Bayesian statistical model for population level modeling of transient light curves, and discuss its implementation using an efficient Hamiltonian Monte Carlo technique. As a test case, we apply this model to the Type IIP SN sample from the Pan-STARRS1 Medium Deep Survey, consisting of 18,837 photometric observations of 76 SNe, corresponding to a joint posterior distribution with 9176 parameters under our model. Our hierarchical model fits provide improved constraints on light curve parameters relevant to the physical properties of their progenitor stars relative to modeling individual light curves alone. Moreover, we directly evaluate the probability for occurrence rates of unseen light curve characteristics from the model hyperparameters, addressing observational biases in survey methodology. We view this modeling framework as an unsupervised machine learning technique with the ability to maximize scientific returns from data to be collected by future wide field transient searches like LSST

  6. UNSUPERVISED TRANSIENT LIGHT CURVE ANALYSIS VIA HIERARCHICAL BAYESIAN INFERENCE

    Energy Technology Data Exchange (ETDEWEB)

    Sanders, N. E.; Soderberg, A. M. [Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138 (United States); Betancourt, M., E-mail: nsanders@cfa.harvard.edu [Department of Statistics, University of Warwick, Coventry CV4 7AL (United Kingdom)

    2015-02-10

    Historically, light curve studies of supernovae (SNe) and other transient classes have focused on individual objects with copious and high signal-to-noise observations. In the nascent era of wide field transient searches, objects with detailed observations are decreasing as a fraction of the overall known SN population, and this strategy sacrifices the majority of the information contained in the data about the underlying population of transients. A population level modeling approach, simultaneously fitting all available observations of objects in a transient sub-class of interest, fully mines the data to infer the properties of the population and avoids certain systematic biases. We present a novel hierarchical Bayesian statistical model for population level modeling of transient light curves, and discuss its implementation using an efficient Hamiltonian Monte Carlo technique. As a test case, we apply this model to the Type IIP SN sample from the Pan-STARRS1 Medium Deep Survey, consisting of 18,837 photometric observations of 76 SNe, corresponding to a joint posterior distribution with 9176 parameters under our model. Our hierarchical model fits provide improved constraints on light curve parameters relevant to the physical properties of their progenitor stars relative to modeling individual light curves alone. Moreover, we directly evaluate the probability for occurrence rates of unseen light curve characteristics from the model hyperparameters, addressing observational biases in survey methodology. We view this modeling framework as an unsupervised machine learning technique with the ability to maximize scientific returns from data to be collected by future wide field transient searches like LSST.

  7. Data mining with unsupervised clustering using photonic micro-ring resonators

    Science.gov (United States)

    McAulay, Alastair D.

    2013-09-01

    Data is commonly moved through optical fiber in modern data centers and may be stored optically. We propose an optical method of data mining for future data centers to enhance performance. For example, in clustering, a form of unsupervised learning, we propose that parameters corresponding to information in a database are converted from analog values to frequencies, as in the brain's neurons, where similar data will have close frequencies. We describe the Wilson-Cowan model for oscillating neurons. In optics we implement the frequencies with micro ring resonators. Due to the influence of weak coupling, a group of resonators will form clusters of similar frequencies that will indicate the desired parameters having close relations. Fewer clusters are formed as clustering proceeds, which allows the creation of a tree showing topics of importance and their relationships in the database. The tree can be used for instance to target advertising and for planning.

  8. Unsupervised Two-Way Clustering of Metagenomic Sequences

    Directory of Open Access Journals (Sweden)

    Shruthi Prabhakara

    2012-01-01

    Full Text Available A major challenge facing metagenomics is the development of tools for the characterization of functional and taxonomic content of vast amounts of short metagenome reads. The efficacy of clustering methods depends on the number of reads in the dataset, the read length and relative abundances of source genomes in the microbial community. In this paper, we formulate an unsupervised naive Bayes multispecies, multidimensional mixture model for reads from a metagenome. We use the proposed model to cluster metagenomic reads by their species of origin and to characterize the abundance of each species. We model the distribution of word counts along a genome as a Gaussian for shorter, frequent words and as a Poisson for longer words that are rare. We employ either a mixture of Gaussians or mixture of Poissons to model reads within each bin. Further, we handle the high-dimensionality and sparsity associated with the data, by grouping the set of words comprising the reads, resulting in a two-way mixture model. Finally, we demonstrate the accuracy and applicability of this method on simulated and real metagenomes. Our method can accurately cluster reads as short as 100 bps and is robust to varying abundances, divergences and read lengths.

  9. Unsupervised spike sorting based on discriminative subspace learning.

    Science.gov (United States)

    Keshtkaran, Mohammad Reza; Yang, Zhi

    2014-01-01

    Spike sorting is a fundamental preprocessing step for many neuroscience studies which rely on the analysis of spike trains. In this paper, we present two unsupervised spike sorting algorithms based on discriminative subspace learning. The first algorithm simultaneously learns the discriminative feature subspace and performs clustering. It uses histogram of features in the most discriminative projection to detect the number of neurons. The second algorithm performs hierarchical divisive clustering that learns a discriminative 1-dimensional subspace for clustering in each level of the hierarchy until achieving almost unimodal distribution in the subspace. The algorithms are tested on synthetic and in-vivo data, and are compared against two widely used spike sorting methods. The comparative results demonstrate that our spike sorting methods can achieve substantially higher accuracy in lower dimensional feature space, and they are highly robust to noise. Moreover, they provide significantly better cluster separability in the learned subspace than in the subspace obtained by principal component analysis or wavelet transform.

  10. Hierarchical clustering using correlation metric and spatial continuity constraint

    Science.gov (United States)

    Stork, Christopher L.; Brewer, Luke N.

    2012-10-02

    Large data sets are analyzed by hierarchical clustering using correlation as a similarity measure. This provides results that are superior to those obtained using a Euclidean distance similarity measure. A spatial continuity constraint may be applied in hierarchical clustering analysis of images.

  11. Parallel Implementation of the Recursive Approximation of an Unsupervised Hierarchical Segmentation Algorithm. Chapter 5

    Science.gov (United States)

    Tilton, James C.; Plaza, Antonio J. (Editor); Chang, Chein-I. (Editor)

    2008-01-01

    The hierarchical image segmentation algorithm (referred to as HSEG) is a hybrid of hierarchical step-wise optimization (HSWO) and constrained spectral clustering that produces a hierarchical set of image segmentations. HSWO is an iterative approach to region grooving segmentation in which the optimal image segmentation is found at N(sub R) regions, given a segmentation at N(sub R+1) regions. HSEG's addition of constrained spectral clustering makes it a computationally intensive algorithm, for all but, the smallest of images. To counteract this, a computationally efficient recursive approximation of HSEG (called RHSEG) has been devised. Further improvements in processing speed are obtained through a parallel implementation of RHSEG. This chapter describes this parallel implementation and demonstrates its computational efficiency on a Landsat Thematic Mapper test scene.

  12. Unsupervised color image segmentation using a lattice algebra clustering technique

    Science.gov (United States)

    Urcid, Gonzalo; Ritter, Gerhard X.

    2011-08-01

    In this paper we introduce a lattice algebra clustering technique for segmenting digital images in the Red-Green- Blue (RGB) color space. The proposed technique is a two step procedure. Given an input color image, the first step determines the finite set of its extreme pixel vectors within the color cube by means of the scaled min-W and max-M lattice auto-associative memory matrices, including the minimum and maximum vector bounds. In the second step, maximal rectangular boxes enclosing each extreme color pixel are found using the Chebychev distance between color pixels; afterwards, clustering is performed by assigning each image pixel to its corresponding maximal box. The two steps in our proposed method are completely unsupervised or autonomous. Illustrative examples are provided to demonstrate the color segmentation results including a brief numerical comparison with two other non-maximal variations of the same clustering technique.

  13. Cluster Based Hierarchical Routing Protocol for Wireless Sensor Network

    OpenAIRE

    Rashed, Md. Golam; Kabir, M. Hasnat; Rahim, Muhammad Sajjadur; Ullah, Shaikh Enayet

    2012-01-01

    The efficient use of energy source in a sensor node is most desirable criteria for prolong the life time of wireless sensor network. In this paper, we propose a two layer hierarchical routing protocol called Cluster Based Hierarchical Routing Protocol (CBHRP). We introduce a new concept called head-set, consists of one active cluster head and some other associate cluster heads within a cluster. The head-set members are responsible for control and management of the network. Results show that t...

  14. Clinical fracture risk evaluated by hierarchical agglomerative clustering

    DEFF Research Database (Denmark)

    Kruse, C; Eiken, P; Vestergaard, P

    2017-01-01

    reimbursement, primary healthcare sector use and comorbidity of female subjects were combined. Standardized variable means, Euclidean distances and Ward's D2 method of hierarchical agglomerative clustering (HAC), were used to form the clustering object. K number of clusters was selected with the lowest cluster...

  15. Unsupervised learning algorithms

    CERN Document Server

    Aydin, Kemal

    2016-01-01

    This book summarizes the state-of-the-art in unsupervised learning. The contributors discuss how with the proliferation of massive amounts of unlabeled data, unsupervised learning algorithms, which can automatically discover interesting and useful patterns in such data, have gained popularity among researchers and practitioners. The authors outline how these algorithms have found numerous applications including pattern recognition, market basket analysis, web mining, social network analysis, information retrieval, recommender systems, market research, intrusion detection, and fraud detection. They present how the difficulty of developing theoretically sound approaches that are amenable to objective evaluation have resulted in the proposal of numerous unsupervised learning algorithms over the past half-century. The intended audience includes researchers and practitioners who are increasingly using unsupervised learning algorithms to analyze their data. Topics of interest include anomaly detection, clustering,...

  16. Merging K-means with hierarchical clustering for identifying general-shaped groups.

    Science.gov (United States)

    Peterson, Anna D; Ghosh, Arka P; Maitra, Ranjan

    2018-01-01

    Clustering partitions a dataset such that observations placed together in a group are similar but different from those in other groups. Hierarchical and K -means clustering are two approaches but have different strengths and weaknesses. For instance, hierarchical clustering identifies groups in a tree-like structure but suffers from computational complexity in large datasets while K -means clustering is efficient but designed to identify homogeneous spherically-shaped clusters. We present a hybrid non-parametric clustering approach that amalgamates the two methods to identify general-shaped clusters and that can be applied to larger datasets. Specifically, we first partition the dataset into spherical groups using K -means. We next merge these groups using hierarchical methods with a data-driven distance measure as a stopping criterion. Our proposal has the potential to reveal groups with general shapes and structure in a dataset. We demonstrate good performance on several simulated and real datasets.

  17. Robust Pseudo-Hierarchical Support Vector Clustering

    DEFF Research Database (Denmark)

    Hansen, Michael Sass; Sjöstrand, Karl; Olafsdóttir, Hildur

    2007-01-01

    Support vector clustering (SVC) has proven an efficient algorithm for clustering of noisy and high-dimensional data sets, with applications within many fields of research. An inherent problem, however, has been setting the parameters of the SVC algorithm. Using the recent emergence of a method...... for calculating the entire regularization path of the support vector domain description, we propose a fast method for robust pseudo-hierarchical support vector clustering (HSVC). The method is demonstrated to work well on generated data, as well as for detecting ischemic segments from multidimensional myocardial...

  18. Hierarchical modeling of cluster size in wildlife surveys

    Science.gov (United States)

    Royle, J. Andrew

    2008-01-01

    Clusters or groups of individuals are the fundamental unit of observation in many wildlife sampling problems, including aerial surveys of waterfowl, marine mammals, and ungulates. Explicit accounting of cluster size in models for estimating abundance is necessary because detection of individuals within clusters is not independent and detectability of clusters is likely to increase with cluster size. This induces a cluster size bias in which the average cluster size in the sample is larger than in the population at large. Thus, failure to account for the relationship between delectability and cluster size will tend to yield a positive bias in estimates of abundance or density. I describe a hierarchical modeling framework for accounting for cluster-size bias in animal sampling. The hierarchical model consists of models for the observation process conditional on the cluster size distribution and the cluster size distribution conditional on the total number of clusters. Optionally, a spatial model can be specified that describes variation in the total number of clusters per sample unit. Parameter estimation, model selection, and criticism may be carried out using conventional likelihood-based methods. An extension of the model is described for the situation where measurable covariates at the level of the sample unit are available. Several candidate models within the proposed class are evaluated for aerial survey data on mallard ducks (Anas platyrhynchos).

  19. Classification of behavior using unsupervised temporal neural networks

    International Nuclear Information System (INIS)

    Adair, K.L.

    1998-03-01

    Adding recurrent connections to unsupervised neural networks used for clustering creates a temporal neural network which clusters a sequence of inputs as they appear over time. The model presented combines the Jordan architecture with the unsupervised learning technique Adaptive Resonance Theory, Fuzzy ART. The combination yields a neural network capable of quickly clustering sequential pattern sequences as the sequences are generated. The applicability of the architecture is illustrated through a facility monitoring problem

  20. Unsupervised Feature Subset Selection

    DEFF Research Database (Denmark)

    Søndberg-Madsen, Nicolaj; Thomsen, C.; Pena, Jose

    2003-01-01

    This paper studies filter and hybrid filter-wrapper feature subset selection for unsupervised learning (data clustering). We constrain the search for the best feature subset by scoring the dependence of every feature on the rest of the features, conjecturing that these scores discriminate some ir...... irrelevant features. We report experimental results on artificial and real data for unsupervised learning of naive Bayes models. Both the filter and hybrid approaches perform satisfactorily....

  1. Unsupervised classification of multivariate geostatistical data: Two algorithms

    Science.gov (United States)

    Romary, Thomas; Ors, Fabien; Rivoirard, Jacques; Deraisme, Jacques

    2015-12-01

    With the increasing development of remote sensing platforms and the evolution of sampling facilities in mining and oil industry, spatial datasets are becoming increasingly large, inform a growing number of variables and cover wider and wider areas. Therefore, it is often necessary to split the domain of study to account for radically different behaviors of the natural phenomenon over the domain and to simplify the subsequent modeling step. The definition of these areas can be seen as a problem of unsupervised classification, or clustering, where we try to divide the domain into homogeneous domains with respect to the values taken by the variables in hand. The application of classical clustering methods, designed for independent observations, does not ensure the spatial coherence of the resulting classes. Image segmentation methods, based on e.g. Markov random fields, are not adapted to irregularly sampled data. Other existing approaches, based on mixtures of Gaussian random functions estimated via the expectation-maximization algorithm, are limited to reasonable sample sizes and a small number of variables. In this work, we propose two algorithms based on adaptations of classical algorithms to multivariate geostatistical data. Both algorithms are model free and can handle large volumes of multivariate, irregularly spaced data. The first one proceeds by agglomerative hierarchical clustering. The spatial coherence is ensured by a proximity condition imposed for two clusters to merge. This proximity condition relies on a graph organizing the data in the coordinates space. The hierarchical algorithm can then be seen as a graph-partitioning algorithm. Following this interpretation, a spatial version of the spectral clustering algorithm is also proposed. The performances of both algorithms are assessed on toy examples and a mining dataset.

  2. Energy Efficient Hierarchical Clustering Approaches in Wireless Sensor Networks: A Survey

    Directory of Open Access Journals (Sweden)

    Bilal Jan

    2017-01-01

    Full Text Available Wireless sensor networks (WSN are one of the significant technologies due to their diverse applications such as health care monitoring, smart phones, military, disaster management, and other surveillance systems. Sensor nodes are usually deployed in large number that work independently in unattended harsh environments. Due to constraint resources, typically the scarce battery power, these wireless nodes are grouped into clusters for energy efficient communication. In clustering hierarchical schemes have achieved great interest for minimizing energy consumption. Hierarchical schemes are generally categorized as cluster-based and grid-based approaches. In cluster-based approaches, nodes are grouped into clusters, where a resourceful sensor node is nominated as a cluster head (CH while in grid-based approach the network is divided into confined virtual grids usually performed by the base station. This paper highlights and discusses the design challenges for cluster-based schemes, the important cluster formation parameters, and classification of hierarchical clustering protocols. Moreover, existing cluster-based and grid-based techniques are evaluated by considering certain parameters to help users in selecting appropriate technique. Furthermore, a detailed summary of these protocols is presented with their advantages, disadvantages, and applicability in particular cases.

  3. Clustering-based classification of road traffic accidents using hierarchical clustering and artificial neural networks.

    Science.gov (United States)

    Taamneh, Madhar; Taamneh, Salah; Alkheder, Sharaf

    2017-09-01

    Artificial neural networks (ANNs) have been widely used in predicting the severity of road traffic crashes. All available information about previously occurred accidents is typically used for building a single prediction model (i.e., classifier). Too little attention has been paid to the differences between these accidents, leading, in most cases, to build less accurate predictors. Hierarchical clustering is a well-known clustering method that seeks to group data by creating a hierarchy of clusters. Using hierarchical clustering and ANNs, a clustering-based classification approach for predicting the injury severity of road traffic accidents was proposed. About 6000 road accidents occurred over a six-year period from 2008 to 2013 in Abu Dhabi were used throughout this study. In order to reduce the amount of variation in data, hierarchical clustering was applied on the data set to organize it into six different forms, each with different number of clusters (i.e., clusters from 1 to 6). Two ANN models were subsequently built for each cluster of accidents in each generated form. The first model was built and validated using all accidents (training set), whereas only 66% of the accidents were used to build the second model, and the remaining 34% were used to test it (percentage split). Finally, the weighted average accuracy was computed for each type of models in each from of data. The results show that when testing the models using the training set, clustering prior to classification achieves (11%-16%) more accuracy than without using clustering, while the percentage split achieves (2%-5%) more accuracy. The results also suggest that partitioning the accidents into six clusters achieves the best accuracy if both types of models are taken into account.

  4. Unsupervised Performance Evaluation Strategy for Bridge Superstructure Based on Fuzzy Clustering and Field Data

    Directory of Open Access Journals (Sweden)

    Yubo Jiao

    2013-01-01

    Full Text Available Performance evaluation of a bridge is critical for determining the optimal maintenance strategy. An unsupervised bridge superstructure state assessment method is proposed in this paper based on fuzzy clustering and bridge field measured data. Firstly, the evaluation index system of bridge is constructed. Secondly, a certain number of bridge health monitoring data are selected as clustering samples to obtain the fuzzy similarity matrix and fuzzy equivalent matrix. Finally, different thresholds are selected to form dynamic clustering maps and determine the best classification based on statistic analysis. The clustering result is regarded as a sample base, and the bridge state can be evaluated by calculating the fuzzy nearness between the unknown bridge state data and the sample base. Nanping Bridge in Jilin Province is selected as the engineering project to verify the effectiveness of the proposed method.

  5. The reflection of hierarchical cluster analysis of co-occurrence matrices in SPSS

    NARCIS (Netherlands)

    Zhou, Q.; Leng, F.; Leydesdorff, L.

    2015-01-01

    Purpose: To discuss the problems arising from hierarchical cluster analysis of co-occurrence matrices in SPSS, and the corresponding solutions. Design/methodology/approach: We design different methods of using the SPSS hierarchical clustering module for co-occurrence matrices in order to compare

  6. A Hierarchical Clustering Methodology for the Estimation of Toxicity

    Science.gov (United States)

    A Quantitative Structure Activity Relationship (QSAR) methodology based on hierarchical clustering was developed to predict toxicological endpoints. This methodology utilizes Ward's method to divide a training set into a series of structurally similar clusters. The structural sim...

  7. A Novel Divisive Hierarchical Clustering Algorithm for Geospatial Analysis

    Directory of Open Access Journals (Sweden)

    Shaoning Li

    2017-01-01

    Full Text Available In the fields of geographic information systems (GIS and remote sensing (RS, the clustering algorithm has been widely used for image segmentation, pattern recognition, and cartographic generalization. Although clustering analysis plays a key role in geospatial modelling, traditional clustering methods are limited due to computational complexity, noise resistant ability and robustness. Furthermore, traditional methods are more focused on the adjacent spatial context, which makes it hard for the clustering methods to be applied to multi-density discrete objects. In this paper, a new method, cell-dividing hierarchical clustering (CDHC, is proposed based on convex hull retraction. The main steps are as follows. First, a convex hull structure is constructed to describe the global spatial context of geospatial objects. Then, the retracting structure of each borderline is established in sequence by setting the initial parameter. The objects are split into two clusters (i.e., “sub-clusters” if the retracting structure intersects with the borderlines. Finally, clusters are repeatedly split and the initial parameter is updated until the terminate condition is satisfied. The experimental results show that CDHC separates the multi-density objects from noise sufficiently and also reduces complexity compared to the traditional agglomerative hierarchical clustering algorithm.

  8. ProtoBee: Hierarchical classification and annotation of the honey bee proteome

    OpenAIRE

    Kaplan, Noam; Linial, Michal

    2006-01-01

    The recently sequenced genome of the honey bee (Apis mellifera) has produced 10,157 predicted protein sequences, calling for a computational effort to extract biological insights from them. We have applied an unsupervised hierarchical protein-clustering method, which was previously used in the ProtoNet system, to nearly 200,000 proteins consisting of the predicted honey bee proteins, the SWISS-PROT protein database, and the complete set of proteins of the mouse (Mus musculus) and the fruit fl...

  9. Hierarchical Control for Multiple DC Microgrids Clusters

    DEFF Research Database (Denmark)

    Shafiee, Qobad; Dragicevic, Tomislav; Vasquez, Juan Carlos

    2014-01-01

    This paper presents a distributed hierarchical control framework to ensure reliable operation of dc Microgrid (MG) clusters. In this hierarchy, primary control is used to regulate the common bus voltage inside each MG locally. An adaptive droop method is proposed for this level which determines...

  10. Hierarchical video summarization based on context clustering

    Science.gov (United States)

    Tseng, Belle L.; Smith, John R.

    2003-11-01

    A personalized video summary is dynamically generated in our video personalization and summarization system based on user preference and usage environment. The three-tier personalization system adopts the server-middleware-client architecture in order to maintain, select, adapt, and deliver rich media content to the user. The server stores the content sources along with their corresponding MPEG-7 metadata descriptions. In this paper, the metadata includes visual semantic annotations and automatic speech transcriptions. Our personalization and summarization engine in the middleware selects the optimal set of desired video segments by matching shot annotations and sentence transcripts with user preferences. Besides finding the desired contents, the objective is to present a coherent summary. There are diverse methods for creating summaries, and we focus on the challenges of generating a hierarchical video summary based on context information. In our summarization algorithm, three inputs are used to generate the hierarchical video summary output. These inputs are (1) MPEG-7 metadata descriptions of the contents in the server, (2) user preference and usage environment declarations from the user client, and (3) context information including MPEG-7 controlled term list and classification scheme. In a video sequence, descriptions and relevance scores are assigned to each shot. Based on these shot descriptions, context clustering is performed to collect consecutively similar shots to correspond to hierarchical scene representations. The context clustering is based on the available context information, and may be derived from domain knowledge or rules engines. Finally, the selection of structured video segments to generate the hierarchical summary efficiently balances between scene representation and shot selection.

  11. Data clustering theory, algorithms, and applications

    CERN Document Server

    Gan, Guojun; Wu, Jianhong

    2007-01-01

    Cluster analysis is an unsupervised process that divides a set of objects into homogeneous groups. This book starts with basic information on cluster analysis, including the classification of data and the corresponding similarity measures, followed by the presentation of over 50 clustering algorithms in groups according to some specific baseline methodologies such as hierarchical, center-based, and search-based methods. As a result, readers and users can easily identify an appropriate algorithm for their applications and compare novel ideas with existing results. The book also provides examples of clustering applications to illustrate the advantages and shortcomings of different clustering architectures and algorithms. Application areas include pattern recognition, artificial intelligence, information technology, image processing, biology, psychology, and marketing. Readers also learn how to perform cluster analysis with the C/C++ and MATLAB® programming languages.

  12. Comparative analysis of clustering methods for gene expression time course data

    Directory of Open Access Journals (Sweden)

    Ivan G. Costa

    2004-01-01

    Full Text Available This work performs a data driven comparative study of clustering methods used in the analysis of gene expression time courses (or time series. Five clustering methods found in the literature of gene expression analysis are compared: agglomerative hierarchical clustering, CLICK, dynamical clustering, k-means and self-organizing maps. In order to evaluate the methods, a k-fold cross-validation procedure adapted to unsupervised methods is applied. The accuracy of the results is assessed by the comparison of the partitions obtained in these experiments with gene annotation, such as protein function and series classification.

  13. Decomposition methods for unsupervised learning

    DEFF Research Database (Denmark)

    Mørup, Morten

    2008-01-01

    This thesis presents the application and development of decomposition methods for Unsupervised Learning. It covers topics from classical factor analysis based decomposition and its variants such as Independent Component Analysis, Non-negative Matrix Factorization and Sparse Coding...... methods and clustering problems is derived both in terms of classical point clustering but also in terms of community detection in complex networks. A guiding principle throughout this thesis is the principle of parsimony. Hence, the goal of Unsupervised Learning is here posed as striving for simplicity...... in the decompositions. Thus, it is demonstrated how a wide range of decomposition methods explicitly or implicitly strive to attain this goal. Applications of the derived decompositions are given ranging from multi-media analysis of image and sound data, analysis of biomedical data such as electroencephalography...

  14. Unsupervised Learning and Generalization

    DEFF Research Database (Denmark)

    Hansen, Lars Kai; Larsen, Jan

    1996-01-01

    The concept of generalization is defined for a general class of unsupervised learning machines. The generalization error is a straightforward extension of the corresponding concept for supervised learning, and may be estimated empirically using a test set or by statistical means-in close analogy ...... with supervised learning. The empirical and analytical estimates are compared for principal component analysis and for K-means clustering based density estimation......The concept of generalization is defined for a general class of unsupervised learning machines. The generalization error is a straightforward extension of the corresponding concept for supervised learning, and may be estimated empirically using a test set or by statistical means-in close analogy...

  15. Improved Gravitation Field Algorithm and Its Application in Hierarchical Clustering

    Science.gov (United States)

    Zheng, Ming; Sun, Ying; Liu, Gui-xia; Zhou, You; Zhou, Chun-guang

    2012-01-01

    Background Gravitation field algorithm (GFA) is a new optimization algorithm which is based on an imitation of natural phenomena. GFA can do well both for searching global minimum and multi-minima in computational biology. But GFA needs to be improved for increasing efficiency, and modified for applying to some discrete data problems in system biology. Method An improved GFA called IGFA was proposed in this paper. Two parts were improved in IGFA. The first one is the rule of random division, which is a reasonable strategy and makes running time shorter. The other one is rotation factor, which can improve the accuracy of IGFA. And to apply IGFA to the hierarchical clustering, the initial part and the movement operator were modified. Results Two kinds of experiments were used to test IGFA. And IGFA was applied to hierarchical clustering. The global minimum experiment was used with IGFA, GFA, GA (genetic algorithm) and SA (simulated annealing). Multi-minima experiment was used with IGFA and GFA. The two experiments results were compared with each other and proved the efficiency of IGFA. IGFA is better than GFA both in accuracy and running time. For the hierarchical clustering, IGFA is used to optimize the smallest distance of genes pairs, and the results were compared with GA and SA, singular-linkage clustering, UPGMA. The efficiency of IGFA is proved. PMID:23173043

  16. The clustering-based case-based reasoning for imbalanced business failure prediction: a hybrid approach through integrating unsupervised process with supervised process

    Science.gov (United States)

    Li, Hui; Yu, Jun-Ling; Yu, Le-An; Sun, Jie

    2014-05-01

    Case-based reasoning (CBR) is one of the main forecasting methods in business forecasting, which performs well in prediction and holds the ability of giving explanations for the results. In business failure prediction (BFP), the number of failed enterprises is relatively small, compared with the number of non-failed ones. However, the loss is huge when an enterprise fails. Therefore, it is necessary to develop methods (trained on imbalanced samples) which forecast well for this small proportion of failed enterprises and performs accurately on total accuracy meanwhile. Commonly used methods constructed on the assumption of balanced samples do not perform well in predicting minority samples on imbalanced samples consisting of the minority/failed enterprises and the majority/non-failed ones. This article develops a new method called clustering-based CBR (CBCBR), which integrates clustering analysis, an unsupervised process, with CBR, a supervised process, to enhance the efficiency of retrieving information from both minority and majority in CBR. In CBCBR, various case classes are firstly generated through hierarchical clustering inside stored experienced cases, and class centres are calculated out by integrating cases information in the same clustered class. When predicting the label of a target case, its nearest clustered case class is firstly retrieved by ranking similarities between the target case and each clustered case class centre. Then, nearest neighbours of the target case in the determined clustered case class are retrieved. Finally, labels of the nearest experienced cases are used in prediction. In the empirical experiment with two imbalanced samples from China, the performance of CBCBR was compared with the classical CBR, a support vector machine, a logistic regression and a multi-variant discriminate analysis. The results show that compared with the other four methods, CBCBR performed significantly better in terms of sensitivity for identifying the

  17. Hierarchical clusters of phytoplankton variables in dammed water bodies

    Science.gov (United States)

    Silva, Eliana Costa e.; Lopes, Isabel Cristina; Correia, Aldina; Gonçalves, A. Manuela

    2017-06-01

    In this paper a dataset containing biological variables of the water column of several Portuguese reservoirs is analyzed. Hierarchical cluster analysis is used to obtain clusters of phytoplankton variables of the phylum Cyanophyta, with the objective of validating the classification of Portuguese reservoirs previewly presented in [1] which were divided into three clusters: (1) Interior Tagus and Aguieira; (2) Douro; and (3) Other rivers. Now three new clusters of Cyanophyta variables were found. Kruskal-Wallis and Mann-Whitney tests are used to compare the now obtained Cyanophyta clusters and the previous Reservoirs clusters, in order to validate the classification of the water quality of reservoirs. The amount of Cyanophyta algae present in the reservoirs from the three clusters is significantly different, which validates the previous classification.

  18. An Improved Unsupervised Modeling Methodology For Detecting Fraud In Vendor Payment Transactions

    National Research Council Canada - National Science Library

    Rouillard, Gregory

    2003-01-01

    ...) vendor payment transactions through Unsupervised Modeling (cluster analysis) . Clementine Data Mining software is used to construct unsupervised models of vendor payment data using the K-Means, Two Step, and Kohonen algorithms...

  19. A hierarchical cluster analysis of normal-tension glaucoma using spectral-domain optical coherence tomography parameters.

    Science.gov (United States)

    Bae, Hyoung Won; Ji, Yongwoo; Lee, Hye Sun; Lee, Naeun; Hong, Samin; Seong, Gong Je; Sung, Kyung Rim; Kim, Chan Yun

    2015-01-01

    Normal-tension glaucoma (NTG) is a heterogenous disease, and there is still controversy about subclassifications of this disorder. On the basis of spectral-domain optical coherence tomography (SD-OCT), we subdivided NTG with hierarchical cluster analysis using optic nerve head (ONH) parameters and retinal nerve fiber layer (RNFL) thicknesses. A total of 200 eyes of 200 NTG patients between March 2011 and June 2012 underwent SD-OCT scans to measure ONH parameters and RNFL thicknesses. We classified NTG into homogenous subgroups based on these variables using a hierarchical cluster analysis, and compared clusters to evaluate diverse NTG characteristics. Three clusters were found after hierarchical cluster analysis. Cluster 1 (62 eyes) had the thickest RNFL and widest rim area, and showed early glaucoma features. Cluster 2 (60 eyes) was characterized by the largest cup/disc ratio and cup volume, and showed advanced glaucomatous damage. Cluster 3 (78 eyes) had small disc areas in SD-OCT and were comprised of patients with significantly younger age, longer axial length, and greater myopia than the other 2 groups. A hierarchical cluster analysis of SD-OCT scans divided NTG patients into 3 groups based upon ONH parameters and RNFL thicknesses. It is anticipated that the small disc area group comprised of younger and more myopic patients may show unique features unlike the other 2 groups.

  20. Surface mapping via unsupervised classification of remote sensing: application to MESSENGER/MASCS and DAWN/VIRS data.

    Science.gov (United States)

    D'Amore, M.; Le Scaon, R.; Helbert, J.; Maturilli, A.

    2017-12-01

    Machine-learning achieved unprecedented results in high-dimensional data processing tasks with wide applications in various fields. Due to the growing number of complex nonlinear systems that have to be investigated in science and the bare raw size of data nowadays available, ML offers the unique ability to extract knowledge, regardless the specific application field. Examples are image segmentation, supervised/unsupervised/ semi-supervised classification, feature extraction, data dimensionality analysis/reduction.The MASCS instrument has mapped Mercury surface in the 400-1145 nm wavelength range during orbital observations by the MESSENGER spacecraft. We have conducted k-means unsupervised hierarchical clustering to identify and characterize spectral units from MASCS observations. The results display a dichotomy: a polar and equatorial units, possibly linked to compositional differences or weathering due to irradiation. To explore possible relations between composition and spectral behavior, we have compared the spectral provinces with elemental abundance maps derived from MESSENGER's X-Ray Spectrometer (XRS).For the Vesta application on DAWN Visible and infrared spectrometer (VIR) data, we explored several Machine Learning techniques: image segmentation method, stream algorithm and hierarchical clustering.The algorithm successfully separates the Olivine outcrops around two craters on Vesta's surface [1]. New maps summarizing the spectral and chemical signature of the surface could be automatically produced.We conclude that instead of hand digging in data, scientist could choose a subset of algorithms with well known feature (i.e. efficacy on the particular problem, speed, accuracy) and focus their effort in understanding what important characteristic of the groups found in the data mean. [1] E Ammannito et al. "Olivine in an unexpected location on Vesta's surface". In: Nature 504.7478 (2013), pp. 122-125.

  1. Hierarchical cluster analysis of progression patterns in open-angle glaucoma patients with medical treatment.

    Science.gov (United States)

    Bae, Hyoung Won; Rho, Seungsoo; Lee, Hye Sun; Lee, Naeun; Hong, Samin; Seong, Gong Je; Sung, Kyung Rim; Kim, Chan Yun

    2014-04-29

    To classify medically treated open-angle glaucoma (OAG) by the pattern of progression using hierarchical cluster analysis, and to determine OAG progression characteristics by comparing clusters. Ninety-five eyes of 95 OAG patients who received medical treatment, and who had undergone visual field (VF) testing at least once per year for 5 or more years. OAG was classified into subgroups using hierarchical cluster analysis based on the following five variables: baseline mean deviation (MD), baseline visual field index (VFI), MD slope, VFI slope, and Glaucoma Progression Analysis (GPA) printout. After that, other parameters were compared between clusters. Two clusters were made after a hierarchical cluster analysis. Cluster 1 showed -4.06 ± 2.43 dB baseline MD, 92.58% ± 6.27% baseline VFI, -0.28 ± 0.38 dB per year MD slope, -0.52% ± 0.81% per year VFI slope, and all "no progression" cases in GPA printout, whereas cluster 2 showed -8.68 ± 3.81 baseline MD, 77.54 ± 12.98 baseline VFI, -0.72 ± 0.55 MD slope, -2.22 ± 1.89 VFI slope, and seven "possible" and four "likely" progression cases in GPA printout. There were no significant differences in age, sex, mean IOP, central corneal thickness, and axial length between clusters. However, cluster 2 included more high-tension glaucoma patients and used a greater number of antiglaucoma eye drops significantly compared with cluster 1. Hierarchical cluster analysis of progression patterns divided OAG into slow and fast progression groups, evidenced by assessing the parameters of glaucomatous progression in VF testing. In the fast progression group, the prevalence of high-tension glaucoma was greater and the number of antiglaucoma medications administered was increased versus the slow progression group. Copyright 2014 The Association for Research in Vision and Ophthalmology, Inc.

  2. Analysis of genetic association using hierarchical clustering and cluster validation indices.

    Science.gov (United States)

    Pagnuco, Inti A; Pastore, Juan I; Abras, Guillermo; Brun, Marcel; Ballarin, Virginia L

    2017-10-01

    It is usually assumed that co-expressed genes suggest co-regulation in the underlying regulatory network. Determining sets of co-expressed genes is an important task, based on some criteria of similarity. This task is usually performed by clustering algorithms, where the genes are clustered into meaningful groups based on their expression values in a set of experiment. In this work, we propose a method to find sets of co-expressed genes, based on cluster validation indices as a measure of similarity for individual gene groups, and a combination of variants of hierarchical clustering to generate the candidate groups. We evaluated its ability to retrieve significant sets on simulated correlated and real genomics data, where the performance is measured based on its detection ability of co-regulated sets against a full search. Additionally, we analyzed the quality of the best ranked groups using an online bioinformatics tool that provides network information for the selected genes. Copyright © 2017 Elsevier Inc. All rights reserved.

  3. Star Cluster Structure from Hierarchical Star Formation

    Science.gov (United States)

    Grudic, Michael; Hopkins, Philip; Murray, Norman; Lamberts, Astrid; Guszejnov, David; Schmitz, Denise; Boylan-Kolchin, Michael

    2018-01-01

    Young massive star clusters (YMCs) spanning 104-108 M⊙ in mass generally have similar radial surface density profiles, with an outer power-law index typically between -2 and -3. This similarity suggests that they are shaped by scale-free physics at formation. Recent multi-physics MHD simulations of YMC formation have also produced populations of YMCs with this type of surface density profile, allowing us to narrow down the physics necessary to form a YMC with properties as observed. We show that the shallow density profiles of YMCs are a natural result of phase-space mixing that occurs as they assemble from the clumpy, hierarchically-clustered configuration imprinted by the star formation process. We develop physical intuition for this process via analytic arguments and collisionless N-body experiments, elucidating the connection between star formation physics and star cluster structure. This has implications for the early-time structure and evolution of proto-globular clusters, and prospects for simulating their formation in the FIRE cosmological zoom-in simulations.

  4. Down-Regulation of Olfactory Receptors in Response to Traumatic Brain Injury Promotes Risk for Alzheimers Disease

    Science.gov (United States)

    2015-12-01

    group assignment of samples in unsupervised hierarchical clustering by the Unweighted Pair-Group Method using Arithmetic averages ( UPGMA ) based on...log2 transformed MAS5.0 signal values; probe set clustering was performed by the UPGMA method using Cosine correlation as the similarity met- ric. For...differentially-regulated genes identified were subjected to unsupervised hierarchical clustering analysis using the UPGMA algorithm with cosine correlation as

  5. Down-Regulation of Olfactory Receptors in Response to Traumatic Brain Injury Promotes Risk for Alzheimer’s Disease

    Science.gov (United States)

    2013-10-01

    correct group assignment of samples in unsupervised hierarchical clustering by the Unweighted Pair-Group Method using Arithmetic averages ( UPGMA ) based on...centering of log2 transformed MAS5.0 signal values; probe set clustering was performed by the UPGMA method using Cosine correlation as the similarity met...A) The 108 differentially-regulated genes identified were subjected to unsupervised hierarchical clustering analysis using the UPGMA algorithm with

  6. The structure of nearby clusters of galaxies Hierarchical clustering and an application to the Leo region

    CERN Document Server

    Materne, J

    1978-01-01

    A new method of classifying groups of galaxies, called hierarchical clustering, is presented as a tool for the investigation of nearby groups of galaxies. The method is free from model assumptions about the groups. The scaling of the different coordinates is necessary, and the level from which one accepts the groups as real has to be determined. Hierarchical clustering is applied to an unbiased sample of galaxies in the Leo region. Five distinct groups result which have reasonable physical properties, such as low crossing times and conservative mass-to-light ratios, and which follow a radial velocity- luminosity relation. Only 4 out of 39 galaxies were adopted as field galaxies. (27 refs).

  7. The Hierarchical Distribution of the Young Stellar Clusters in Six Local Star-forming Galaxies

    Energy Technology Data Exchange (ETDEWEB)

    Grasha, K.; Calzetti, D. [Astronomy Department, University of Massachusetts, Amherst, MA 01003 (United States); Adamo, A.; Messa, M. [Dept. of Astronomy, The Oskar Klein Centre, Stockholm University, Stockholm (Sweden); Kim, H. [Gemini Observatory, La Serena (Chile); Elmegreen, B. G. [IBM Research Division, T.J. Watson Research Center, Yorktown Hts., NY (United States); Gouliermis, D. A. [Zentrum für Astronomie der Universität Heidelberg, Institut für Theoretische Astrophysik, Albert-Ueberle-Str. 2, D-69120 Heidelberg (Germany); Dale, D. A. [Dept. of Physics and Astronomy, University of Wyoming, Laramie, WY (United States); Fumagalli, M. [Institute for Computational Cosmology and Centre for Extragalactic Astronomy, Durham University, Durham (United Kingdom); Grebel, E. K.; Shabani, F. [Astronomisches Rechen-Institut, Zentrum für Astronomie der Universität Heidelberg, Mönchhofstr. 12-14, D-69120 Heidelberg (Germany); Johnson, K. E. [Dept. of Astronomy, University of Virginia, Charlottesville, VA (United States); Kahre, L. [Dept. of Astronomy, New Mexico State University, Las Cruces, NM (United States); Kennicutt, R. C. [Institute of Astronomy, University of Cambridge, Cambridge (United Kingdom); Pellerin, A. [Dept. of Physics and Astronomy, State University of New York at Geneseo, Geneseo NY (United States); Ryon, J. E.; Ubeda, L. [Space Telescope Science Institute, Baltimore, MD (United States); Smith, L. J. [European Space Agency/Space Telescope Science Institute, Baltimore, MD (United States); Thilker, D., E-mail: kgrasha@astro.umass.edu [Dept. of Physics and Astronomy, The Johns Hopkins University, Baltimore, MD (United States)

    2017-05-10

    We present a study of the hierarchical clustering of the young stellar clusters in six local (3–15 Mpc) star-forming galaxies using Hubble Space Telescope broadband WFC3/UVIS UV and optical images from the Treasury Program LEGUS (Legacy ExtraGalactic UV Survey). We identified 3685 likely clusters and associations, each visually classified by their morphology, and we use the angular two-point correlation function to study the clustering of these stellar systems. We find that the spatial distribution of the young clusters and associations are clustered with respect to each other, forming large, unbound hierarchical star-forming complexes that are in general very young. The strength of the clustering decreases with increasing age of the star clusters and stellar associations, becoming more homogeneously distributed after ∼40–60 Myr and on scales larger than a few hundred parsecs. In all galaxies, the associations exhibit a global behavior that is distinct and more strongly correlated from compact clusters. Thus, populations of clusters are more evolved than associations in terms of their spatial distribution, traveling significantly from their birth site within a few tens of Myr, whereas associations show evidence of disruption occurring very quickly after their formation. The clustering of the stellar systems resembles that of a turbulent interstellar medium that drives the star formation process, correlating the components in unbound star-forming complexes in a hierarchical manner, dispersing shortly after formation, suggestive of a single, continuous mode of star formation across all galaxies.

  8. The Hierarchical Distribution of the Young Stellar Clusters in Six Local Star-forming Galaxies

    Science.gov (United States)

    Grasha, K.; Calzetti, D.; Adamo, A.; Kim, H.; Elmegreen, B. G.; Gouliermis, D. A.; Dale, D. A.; Fumagalli, M.; Grebel, E. K.; Johnson, K. E.; Kahre, L.; Kennicutt, R. C.; Messa, M.; Pellerin, A.; Ryon, J. E.; Smith, L. J.; Shabani, F.; Thilker, D.; Ubeda, L.

    2017-05-01

    We present a study of the hierarchical clustering of the young stellar clusters in six local (3-15 Mpc) star-forming galaxies using Hubble Space Telescope broadband WFC3/UVIS UV and optical images from the Treasury Program LEGUS (Legacy ExtraGalactic UV Survey). We identified 3685 likely clusters and associations, each visually classified by their morphology, and we use the angular two-point correlation function to study the clustering of these stellar systems. We find that the spatial distribution of the young clusters and associations are clustered with respect to each other, forming large, unbound hierarchical star-forming complexes that are in general very young. The strength of the clustering decreases with increasing age of the star clusters and stellar associations, becoming more homogeneously distributed after ˜40-60 Myr and on scales larger than a few hundred parsecs. In all galaxies, the associations exhibit a global behavior that is distinct and more strongly correlated from compact clusters. Thus, populations of clusters are more evolved than associations in terms of their spatial distribution, traveling significantly from their birth site within a few tens of Myr, whereas associations show evidence of disruption occurring very quickly after their formation. The clustering of the stellar systems resembles that of a turbulent interstellar medium that drives the star formation process, correlating the components in unbound star-forming complexes in a hierarchical manner, dispersing shortly after formation, suggestive of a single, continuous mode of star formation across all galaxies.

  9. The Hierarchical Distribution of the Young Stellar Clusters in Six Local Star-forming Galaxies

    International Nuclear Information System (INIS)

    Grasha, K.; Calzetti, D.; Adamo, A.; Messa, M.; Kim, H.; Elmegreen, B. G.; Gouliermis, D. A.; Dale, D. A.; Fumagalli, M.; Grebel, E. K.; Shabani, F.; Johnson, K. E.; Kahre, L.; Kennicutt, R. C.; Pellerin, A.; Ryon, J. E.; Ubeda, L.; Smith, L. J.; Thilker, D.

    2017-01-01

    We present a study of the hierarchical clustering of the young stellar clusters in six local (3–15 Mpc) star-forming galaxies using Hubble Space Telescope broadband WFC3/UVIS UV and optical images from the Treasury Program LEGUS (Legacy ExtraGalactic UV Survey). We identified 3685 likely clusters and associations, each visually classified by their morphology, and we use the angular two-point correlation function to study the clustering of these stellar systems. We find that the spatial distribution of the young clusters and associations are clustered with respect to each other, forming large, unbound hierarchical star-forming complexes that are in general very young. The strength of the clustering decreases with increasing age of the star clusters and stellar associations, becoming more homogeneously distributed after ∼40–60 Myr and on scales larger than a few hundred parsecs. In all galaxies, the associations exhibit a global behavior that is distinct and more strongly correlated from compact clusters. Thus, populations of clusters are more evolved than associations in terms of their spatial distribution, traveling significantly from their birth site within a few tens of Myr, whereas associations show evidence of disruption occurring very quickly after their formation. The clustering of the stellar systems resembles that of a turbulent interstellar medium that drives the star formation process, correlating the components in unbound star-forming complexes in a hierarchical manner, dispersing shortly after formation, suggestive of a single, continuous mode of star formation across all galaxies.

  10. A Distributed Algorithm for the Cluster-Based Outlier Detection Using Unsupervised Extreme Learning Machines

    Directory of Open Access Journals (Sweden)

    Xite Wang

    2017-01-01

    Full Text Available Outlier detection is an important data mining task, whose target is to find the abnormal or atypical objects from a given dataset. The techniques for detecting outliers have a lot of applications, such as credit card fraud detection and environment monitoring. Our previous work proposed the Cluster-Based (CB outlier and gave a centralized method using unsupervised extreme learning machines to compute CB outliers. In this paper, we propose a new distributed algorithm for the CB outlier detection (DACB. On the master node, we collect a small number of points from the slave nodes to obtain a threshold. On each slave node, we design a new filtering method that can use the threshold to efficiently speed up the computation. Furthermore, we also propose a ranking method to optimize the order of cluster scanning. At last, the effectiveness and efficiency of the proposed approaches are verified through a plenty of simulation experiments.

  11. Prediction of Solvent Physical Properties using the Hierarchical Clustering Method

    Science.gov (United States)

    Recently a QSAR (Quantitative Structure Activity Relationship) method, the hierarchical clustering method, was developed to estimate acute toxicity values for large, diverse datasets. This methodology has now been applied to the estimate solvent physical properties including sur...

  12. Determination of genetic structure of germplasm collections: are traditional hierarchical clustering methods appropriate for molecular marker data?

    NARCIS (Netherlands)

    Odong, T.L.; Heerwaarden, van J.; Jansen, J.; Hintum, van T.J.L.; Eeuwijk, van F.A.

    2011-01-01

    Despite the availability of newer approaches, traditional hierarchical clustering remains very popular in genetic diversity studies in plants. However, little is known about its suitability for molecular marker data. We studied the performance of traditional hierarchical clustering techniques using

  13. Unsupervised classification of variable stars

    Science.gov (United States)

    Valenzuela, Lucas; Pichara, Karim

    2018-03-01

    During the past 10 years, a considerable amount of effort has been made to develop algorithms for automatic classification of variable stars. That has been primarily achieved by applying machine learning methods to photometric data sets where objects are represented as light curves. Classifiers require training sets to learn the underlying patterns that allow the separation among classes. Unfortunately, building training sets is an expensive process that demands a lot of human efforts. Every time data come from new surveys; the only available training instances are the ones that have a cross-match with previously labelled objects, consequently generating insufficient training sets compared with the large amounts of unlabelled sources. In this work, we present an algorithm that performs unsupervised classification of variable stars, relying only on the similarity among light curves. We tackle the unsupervised classification problem by proposing an untraditional approach. Instead of trying to match classes of stars with clusters found by a clustering algorithm, we propose a query-based method where astronomers can find groups of variable stars ranked by similarity. We also develop a fast similarity function specific for light curves, based on a novel data structure that allows scaling the search over the entire data set of unlabelled objects. Experiments show that our unsupervised model achieves high accuracy in the classification of different types of variable stars and that the proposed algorithm scales up to massive amounts of light curves.

  14. An Efficient Optimization Method for Solving Unsupervised Data Classification Problems

    Directory of Open Access Journals (Sweden)

    Parvaneh Shabanzadeh

    2015-01-01

    Full Text Available Unsupervised data classification (or clustering analysis is one of the most useful tools and a descriptive task in data mining that seeks to classify homogeneous groups of objects based on similarity and is used in many medical disciplines and various applications. In general, there is no single algorithm that is suitable for all types of data, conditions, and applications. Each algorithm has its own advantages, limitations, and deficiencies. Hence, research for novel and effective approaches for unsupervised data classification is still active. In this paper a heuristic algorithm, Biogeography-Based Optimization (BBO algorithm, was adapted for data clustering problems by modifying the main operators of BBO algorithm, which is inspired from the natural biogeography distribution of different species. Similar to other population-based algorithms, BBO algorithm starts with an initial population of candidate solutions to an optimization problem and an objective function that is calculated for them. To evaluate the performance of the proposed algorithm assessment was carried on six medical and real life datasets and was compared with eight well known and recent unsupervised data classification algorithms. Numerical results demonstrate that the proposed evolutionary optimization algorithm is efficient for unsupervised data classification.

  15. An unsupervised text mining method for relation extraction from biomedical literature.

    Directory of Open Access Journals (Sweden)

    Changqin Quan

    Full Text Available The wealth of interaction information provided in biomedical articles motivated the implementation of text mining approaches to automatically extract biomedical relations. This paper presents an unsupervised method based on pattern clustering and sentence parsing to deal with biomedical relation extraction. Pattern clustering algorithm is based on Polynomial Kernel method, which identifies interaction words from unlabeled data; these interaction words are then used in relation extraction between entity pairs. Dependency parsing and phrase structure parsing are combined for relation extraction. Based on the semi-supervised KNN algorithm, we extend the proposed unsupervised approach to a semi-supervised approach by combining pattern clustering, dependency parsing and phrase structure parsing rules. We evaluated the approaches on two different tasks: (1 Protein-protein interactions extraction, and (2 Gene-suicide association extraction. The evaluation of task (1 on the benchmark dataset (AImed corpus showed that our proposed unsupervised approach outperformed three supervised methods. The three supervised methods are rule based, SVM based, and Kernel based separately. The proposed semi-supervised approach is superior to the existing semi-supervised methods. The evaluation on gene-suicide association extraction on a smaller dataset from Genetic Association Database and a larger dataset from publicly available PubMed showed that the proposed unsupervised and semi-supervised methods achieved much higher F-scores than co-occurrence based method.

  16. Technique for fast and efficient hierarchical clustering

    Science.gov (United States)

    Stork, Christopher

    2013-10-08

    A fast and efficient technique for hierarchical clustering of samples in a dataset includes compressing the dataset to reduce a number of variables within each of the samples of the dataset. A nearest neighbor matrix is generated to identify nearest neighbor pairs between the samples based on differences between the variables of the samples. The samples are arranged into a hierarchy that groups the samples based on the nearest neighbor matrix. The hierarchy is rendered to a display to graphically illustrate similarities or differences between the samples.

  17. The Hierarchical Spectral Merger Algorithm: A New Time Series Clustering Procedure

    KAUST Repository

    Euá n, Carolina; Ombao, Hernando; Ortega, Joaquí n

    2018-01-01

    We present a new method for time series clustering which we call the Hierarchical Spectral Merger (HSM) method. This procedure is based on the spectral theory of time series and identifies series that share similar oscillations or waveforms

  18. Deep Unsupervised Learning on a Desktop PC: A Primer for Cognitive Scientists.

    Science.gov (United States)

    Testolin, Alberto; Stoianov, Ivilin; De Filippo De Grazia, Michele; Zorzi, Marco

    2013-01-01

    Deep belief networks hold great promise for the simulation of human cognition because they show how structured and abstract representations may emerge from probabilistic unsupervised learning. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. However, learning in deep networks typically requires big datasets and it can involve millions of connection weights, which implies that simulations on standard computers are unfeasible. Developing realistic, medium-to-large-scale learning models of cognition would therefore seem to require expertise in programing parallel-computing hardware, and this might explain why the use of this promising approach is still largely confined to the machine learning community. Here we show how simulations of deep unsupervised learning can be easily performed on a desktop PC by exploiting the processors of low cost graphic cards (graphic processor units) without any specific programing effort, thanks to the use of high-level programming routines (available in MATLAB or Python). We also show that even an entry-level graphic card can outperform a small high-performance computing cluster in terms of learning time and with no loss of learning quality. We therefore conclude that graphic card implementations pave the way for a widespread use of deep learning among cognitive scientists for modeling cognition and behavior.

  19. Deep unsupervised learning on a desktop PC: A primer for cognitive scientists

    Directory of Open Access Journals (Sweden)

    Alberto eTestolin

    2013-05-01

    Full Text Available Deep belief networks hold great promise for the simulation of human cognition because they show how structured and abstract representations may emerge from probabilistic unsupervised learning. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. However, learning in deep networks typically requires big datasets and it can involve millions of connection weights, which implies that simulations on standard computers are unfeasible. Developing realistic, medium-to-large-scale learning models of cognition would therefore seem to require expertise in programming parallel-computing hardware, and this might explain why the use of this promising approach is still largely confined to the machine learning community. Here we show how simulations of deep unsupervised learning can be easily performed on a desktop PC by exploiting the processors of low-cost graphic cards (GPUs without any specific programming effort, thanks to the use of high-level programming routines (available in MATLAB or Python. We also show that even an entry-level graphic card can outperform a small high-performance computing cluster in terms of learning time and with no loss of learning quality. We therefore conclude that graphic card implementations pave the way for a widespread use of deep learning among cognitive scientists for modeling cognition and behavior.

  20. Deep Unsupervised Learning on a Desktop PC: A Primer for Cognitive Scientists

    Science.gov (United States)

    Testolin, Alberto; Stoianov, Ivilin; De Filippo De Grazia, Michele; Zorzi, Marco

    2013-01-01

    Deep belief networks hold great promise for the simulation of human cognition because they show how structured and abstract representations may emerge from probabilistic unsupervised learning. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. However, learning in deep networks typically requires big datasets and it can involve millions of connection weights, which implies that simulations on standard computers are unfeasible. Developing realistic, medium-to-large-scale learning models of cognition would therefore seem to require expertise in programing parallel-computing hardware, and this might explain why the use of this promising approach is still largely confined to the machine learning community. Here we show how simulations of deep unsupervised learning can be easily performed on a desktop PC by exploiting the processors of low cost graphic cards (graphic processor units) without any specific programing effort, thanks to the use of high-level programming routines (available in MATLAB or Python). We also show that even an entry-level graphic card can outperform a small high-performance computing cluster in terms of learning time and with no loss of learning quality. We therefore conclude that graphic card implementations pave the way for a widespread use of deep learning among cognitive scientists for modeling cognition and behavior. PMID:23653617

  1. The identification of credit card encoders by hierarchical cluster analysis of the jitters of magnetic stripes.

    Science.gov (United States)

    Leung, S C; Fung, W K; Wong, K H

    1999-01-01

    The relative bit density variation graphs of 207 specimen credit cards processed by 12 encoding machines were examined first visually, and then classified by means of hierarchical cluster analysis. Twenty-nine credit cards being treated as 'questioned' samples were tested by way of cluster analysis against 'controls' derived from known encoders. It was found that hierarchical cluster analysis provided a high accuracy of identification with all 29 'questioned' samples classified correctly. On the other hand, although visual comparison of jitter graphs was less discriminating, it was nevertheless capable of giving a reasonably accurate result.

  2. Analysis of precipitation data in Bangladesh through hierarchical clustering and multidimensional scaling

    Science.gov (United States)

    Rahman, Md. Habibur; Matin, M. A.; Salma, Umma

    2017-12-01

    The precipitation patterns of seventeen locations in Bangladesh from 1961 to 2014 were studied using a cluster analysis and metric multidimensional scaling. In doing so, the current research applies four major hierarchical clustering methods to precipitation in conjunction with different dissimilarity measures and metric multidimensional scaling. A variety of clustering algorithms were used to provide multiple clustering dendrograms for a mixture of distance measures. The dendrogram of pre-monsoon rainfall for the seventeen locations formed five clusters. The pre-monsoon precipitation data for the areas of Srimangal and Sylhet were located in two clusters across the combination of five dissimilarity measures and four hierarchical clustering algorithms. The single linkage algorithm with Euclidian and Manhattan distances, the average linkage algorithm with the Minkowski distance, and Ward's linkage algorithm provided similar results with regard to monsoon precipitation. The results of the post-monsoon and winter precipitation data are shown in different types of dendrograms with disparate combinations of sub-clusters. The schematic geometrical representations of the precipitation data using metric multidimensional scaling showed that the post-monsoon rainfall of Cox's Bazar was located far from those of the other locations. The results of a box-and-whisker plot, different clustering techniques, and metric multidimensional scaling indicated that the precipitation behaviour of Srimangal and Sylhet during the pre-monsoon season, Cox's Bazar and Sylhet during the monsoon season, Maijdi Court and Cox's Bazar during the post-monsoon season, and Cox's Bazar and Khulna during the winter differed from those at other locations in Bangladesh.

  3. Unsupervised classification of major depression using functional connectivity MRI.

    Science.gov (United States)

    Zeng, Ling-Li; Shen, Hui; Liu, Li; Hu, Dewen

    2014-04-01

    The current diagnosis of psychiatric disorders including major depressive disorder based largely on self-reported symptoms and clinical signs may be prone to patients' behaviors and psychiatrists' bias. This study aims at developing an unsupervised machine learning approach for the accurate identification of major depression based on single resting-state functional magnetic resonance imaging scans in the absence of clinical information. Twenty-four medication-naive patients with major depression and 29 demographically similar healthy individuals underwent resting-state functional magnetic resonance imaging. We first clustered the voxels within the perigenual cingulate cortex into two subregions, a subgenual region and a pregenual region, according to their distinct resting-state functional connectivity patterns and showed that a maximum margin clustering-based unsupervised machine learning approach extracted sufficient information from the subgenual cingulate functional connectivity map to differentiate depressed patients from healthy controls with a group-level clustering consistency of 92.5% and an individual-level classification consistency of 92.5%. It was also revealed that the subgenual cingulate functional connectivity network with the highest discriminative power primarily included the ventrolateral and ventromedial prefrontal cortex, superior temporal gyri and limbic areas, indicating that these connections may play critical roles in the pathophysiology of major depression. The current study suggests that subgenual cingulate functional connectivity network signatures may provide promising objective biomarkers for the diagnosis of major depression and that maximum margin clustering-based unsupervised machine learning approaches may have the potential to inform clinical practice and aid in research on psychiatric disorders. Copyright © 2013 Wiley Periodicals, Inc.

  4. Assessment of surface water quality using hierarchical cluster analysis

    Directory of Open Access Journals (Sweden)

    Dheeraj Kumar Dabgerwal

    2016-02-01

    Full Text Available This study was carried out to assess the physicochemical quality river Varuna inVaranasi,India. Water samples were collected from 10 sites during January-June 2015. Pearson correlation analysis was used to assess the direction and strength of relationship between physicochemical parameters. Hierarchical Cluster analysis was also performed to determine the sources of pollution in the river Varuna. The result showed quite high value of DO, Nitrate, BOD, COD and Total Alkalinity, above the BIS permissible limit. The results of correlation analysis identified key water parameters as pH, electrical conductivity, total alkalinity and nitrate, which influence the concentration of other water parameters. Cluster analysis identified three major clusters of sampling sites out of total 10 sites, according to the similarity in water quality. This study illustrated the usefulness of correlation and cluster analysis for getting better information about the river water quality.International Journal of Environment Vol. 5 (1 2016,  pp: 32-44

  5. Constructing storyboards based on hierarchical clustering analysis

    Science.gov (United States)

    Hasebe, Satoshi; Sami, Mustafa M.; Muramatsu, Shogo; Kikuchi, Hisakazu

    2005-07-01

    There are growing needs for quick preview of video contents for the purpose of improving accessibility of video archives as well as reducing network traffics. In this paper, a storyboard that contains a user-specified number of keyframes is produced from a given video sequence. It is based on hierarchical cluster analysis of feature vectors that are derived from wavelet coefficients of video frames. Consistent use of extracted feature vectors is the key to avoid a repetition of computationally-intensive parsing of the same video sequence. Experimental results suggest that a significant reduction in computational time is gained by this strategy.

  6. Hierarchical Network Design

    DEFF Research Database (Denmark)

    Thomadsen, Tommy

    2005-01-01

    Communication networks are immensely important today, since both companies and individuals use numerous services that rely on them. This thesis considers the design of hierarchical (communication) networks. Hierarchical networks consist of layers of networks and are well-suited for coping...... with changing and increasing demands. Two-layer networks consist of one backbone network, which interconnects cluster networks. The clusters consist of nodes and links, which connect the nodes. One node in each cluster is a hub node, and the backbone interconnects the hub nodes of each cluster and thus...... the clusters. The design of hierarchical networks involves clustering of nodes, hub selection, and network design, i.e. selection of links and routing of ows. Hierarchical networks have been in use for decades, but integrated design of these networks has only been considered for very special types of networks...

  7. Which, When, and How: Hierarchical Clustering with Human–Machine Cooperation

    Directory of Open Access Journals (Sweden)

    Huanyang Zheng

    2016-12-01

    Full Text Available Human–Machine Cooperations (HMCs can balance the advantages and disadvantages of human computation (accurate but costly and machine computation (cheap but inaccurate. This paper studies HMCs in agglomerative hierarchical clusterings, where the machine can ask the human some questions. The human will return the answers to the machine, and the machine will use these answers to correct errors in its current clustering results. We are interested in the machine’s strategy on handling the question operations, in terms of three problems: (1 Which question should the machine ask? (2 When should the machine ask the question (early or late? (3 How does the machine adjust the clustering result, if the machine’s mistake is found by the human? Based on the insights of these problems, an efficient algorithm is proposed with five implementation variations. Experiments on image clusterings show that the proposed algorithm can improve the clustering accuracy with few question operations.

  8. Robust Real-Time Music Transcription with a Compositional Hierarchical Model.

    Science.gov (United States)

    Pesek, Matevž; Leonardis, Aleš; Marolt, Matija

    2017-01-01

    The paper presents a new compositional hierarchical model for robust music transcription. Its main features are unsupervised learning of a hierarchical representation of input data, transparency, which enables insights into the learned representation, as well as robustness and speed which make it suitable for real-world and real-time use. The model consists of multiple layers, each composed of a number of parts. The hierarchical nature of the model corresponds well to hierarchical structures in music. The parts in lower layers correspond to low-level concepts (e.g. tone partials), while the parts in higher layers combine lower-level representations into more complex concepts (tones, chords). The layers are learned in an unsupervised manner from music signals. Parts in each layer are compositions of parts from previous layers based on statistical co-occurrences as the driving force of the learning process. In the paper, we present the model's structure and compare it to other hierarchical approaches in the field of music information retrieval. We evaluate the model's performance for the multiple fundamental frequency estimation. Finally, we elaborate on extensions of the model towards other music information retrieval tasks.

  9. The efficiency of average linkage hierarchical clustering algorithm associated multi-scale bootstrap resampling in identifying homogeneous precipitation catchments

    Science.gov (United States)

    Chuan, Zun Liang; Ismail, Noriszura; Shinyie, Wendy Ling; Lit Ken, Tan; Fam, Soo-Fen; Senawi, Azlyna; Yusoff, Wan Nur Syahidah Wan

    2018-04-01

    Due to the limited of historical precipitation records, agglomerative hierarchical clustering algorithms widely used to extrapolate information from gauged to ungauged precipitation catchments in yielding a more reliable projection of extreme hydro-meteorological events such as extreme precipitation events. However, identifying the optimum number of homogeneous precipitation catchments accurately based on the dendrogram resulted using agglomerative hierarchical algorithms are very subjective. The main objective of this study is to propose an efficient regionalized algorithm to identify the homogeneous precipitation catchments for non-stationary precipitation time series. The homogeneous precipitation catchments are identified using average linkage hierarchical clustering algorithm associated multi-scale bootstrap resampling, while uncentered correlation coefficient as the similarity measure. The regionalized homogeneous precipitation is consolidated using K-sample Anderson Darling non-parametric test. The analysis result shows the proposed regionalized algorithm performed more better compared to the proposed agglomerative hierarchical clustering algorithm in previous studies.

  10. Semi-supervised and unsupervised extreme learning machines.

    Science.gov (United States)

    Huang, Gao; Song, Shiji; Gupta, Jatinder N D; Wu, Cheng

    2014-12-01

    Extreme learning machines (ELMs) have proven to be efficient and effective learning mechanisms for pattern classification and regression. However, ELMs are primarily applied to supervised learning problems. Only a few existing research papers have used ELMs to explore unlabeled data. In this paper, we extend ELMs for both semi-supervised and unsupervised tasks based on the manifold regularization, thus greatly expanding the applicability of ELMs. The key advantages of the proposed algorithms are as follows: 1) both the semi-supervised ELM (SS-ELM) and the unsupervised ELM (US-ELM) exhibit learning capability and computational efficiency of ELMs; 2) both algorithms naturally handle multiclass classification or multicluster clustering; and 3) both algorithms are inductive and can handle unseen data at test time directly. Moreover, it is shown in this paper that all the supervised, semi-supervised, and unsupervised ELMs can actually be put into a unified framework. This provides new perspectives for understanding the mechanism of random feature mapping, which is the key concept in ELM theory. Empirical study on a wide range of data sets demonstrates that the proposed algorithms are competitive with the state-of-the-art semi-supervised or unsupervised learning algorithms in terms of accuracy and efficiency.

  11. Content Discovery from Composite Audio : An unsupervised approach

    NARCIS (Netherlands)

    Lu, L.

    2009-01-01

    In this thesis, we developed and assessed a novel robust and unsupervised framework for semantic inference from composite audio signals. We focused on the problem of detecting audio scenes and grouping them into meaningful clusters. Our approach addressed all major steps in a general process of

  12. Exploitation of Clustering Techniques in Transactional Healthcare Data

    Directory of Open Access Journals (Sweden)

    Naeem Ahmad Mahoto

    2014-03-01

    Full Text Available Healthcare service centres equipped with electronic health systems have improved their resources as well as treatment processes. The dynamic nature of healthcare data of each individual makes it complex and difficult for physicians to manually mediate them; therefore, automatic techniques are essential to manage the quality and standardization of treatment procedures. Exploratory data analysis, patternanalysis and grouping of data is managed using clustering techniques, which work as an unsupervised classification. A number of healthcare applications are developed that use several data mining techniques for classification, clustering and extracting useful information from healthcare data. The challenging issue in this domain is to select adequate data mining algorithm for optimal results. This paper exploits three different clustering algorithms: DBSCAN (Density-Based Clustering, agglomerative hierarchical and k-means in real transactional healthcare data of diabetic patients (taken as case study to analyse their performance in large and dispersed healthcare data. The best solution of cluster sets among the exploited algorithms is evaluated using clustering quality indexes and is selected to identify the possible subgroups of patients having similar treatment patterns

  13. The Local Maximum Clustering Method and Its Application in Microarray Gene Expression Data Analysis

    Directory of Open Access Journals (Sweden)

    Chen Yidong

    2004-01-01

    Full Text Available An unsupervised data clustering method, called the local maximum clustering (LMC method, is proposed for identifying clusters in experiment data sets based on research interest. A magnitude property is defined according to research purposes, and data sets are clustered around each local maximum of the magnitude property. By properly defining a magnitude property, this method can overcome many difficulties in microarray data clustering such as reduced projection in similarities, noises, and arbitrary gene distribution. To critically evaluate the performance of this clustering method in comparison with other methods, we designed three model data sets with known cluster distributions and applied the LMC method as well as the hierarchic clustering method, the -mean clustering method, and the self-organized map method to these model data sets. The results show that the LMC method produces the most accurate clustering results. As an example of application, we applied the method to cluster the leukemia samples reported in the microarray study of Golub et al. (1999.

  14. Unsupervised Analysis of Array Comparative Genomic Hybridization Data from Early-Onset Colorectal Cancer Reveals Equivalence with Molecular Classification and Phenotypes

    Directory of Open Access Journals (Sweden)

    María Arriba

    2017-01-01

    Full Text Available AIM: To investigate whether chromosomal instability (CIN is associated with tumor phenotypes and/or with global genomic status based on MSI (microsatellite instability and CIMP (CpG island methylator phenotype in early-onset colorectal cancer (EOCRC. METHODS: Taking as a starting point our previous work in which tumors from 60 EOCRC cases (≤45 years at the time of diagnosis were analyzed by array comparative genomic hybridization (aCGH, in the present study we performed an unsupervised hierarchical clustering analysis of those aCGH data in order to unveil possible associations between the CIN profile and the clinical features of the tumors. In addition, we evaluated the MSI and the CIMP statuses of the samples with the aim of investigating a possible relationship between copy number alterations (CNAs and the MSI/CIMP condition in EOCRC. RESULTS: Based on the similarity of the CNAs detected, the unsupervised analysis stratified samples into two main clusters (A, B and four secondary clusters (A1, A2, B3, B4. The different subgroups showed a certain correspondence with the molecular classification of colorectal cancer (CRC, which enabled us to outline an algorithm to categorize tumors according to their CIMP status. Interestingly, each subcluster showed some distinctive clinicopathological features. But more interestingly, the CIN of each subcluster mainly affected particular chromosomes, allowing us to define chromosomal regions more specifically affected depending on the CIMP/MSI status of the samples. CONCLUSIONS: Our findings may provide a basis for a new form of classifying EOCRC according to the genomic status of the tumors.

  15. flowPeaks: a fast unsupervised clustering for flow cytometry data via K-means and density peak finding.

    Science.gov (United States)

    Ge, Yongchao; Sealfon, Stuart C

    2012-08-01

    For flow cytometry data, there are two common approaches to the unsupervised clustering problem: one is based on the finite mixture model and the other on spatial exploration of the histograms. The former is computationally slow and has difficulty to identify clusters of irregular shapes. The latter approach cannot be applied directly to high-dimensional data as the computational time and memory become unmanageable and the estimated histogram is unreliable. An algorithm without these two problems would be very useful. In this article, we combine ideas from the finite mixture model and histogram spatial exploration. This new algorithm, which we call flowPeaks, can be applied directly to high-dimensional data and identify irregular shape clusters. The algorithm first uses K-means algorithm with a large K to partition the cell population into many small clusters. These partitioned data allow the generation of a smoothed density function using the finite mixture model. All local peaks are exhaustively searched by exploring the density function and the cells are clustered by the associated local peak. The algorithm flowPeaks is automatic, fast and reliable and robust to cluster shape and outliers. This algorithm has been applied to flow cytometry data and it has been compared with state of the art algorithms, including Misty Mountain, FLOCK, flowMeans, flowMerge and FLAME. The R package flowPeaks is available at https://github.com/yongchao/flowPeaks. yongchao.ge@mssm.edu Supplementary data are available at Bioinformatics online.

  16. Segmentation of fluorescence microscopy cell images using unsupervised mining.

    Science.gov (United States)

    Du, Xian; Dua, Sumeet

    2010-05-28

    The accurate measurement of cell and nuclei contours are critical for the sensitive and specific detection of changes in normal cells in several medical informatics disciplines. Within microscopy, this task is facilitated using fluorescence cell stains, and segmentation is often the first step in such approaches. Due to the complex nature of cell issues and problems inherent to microscopy, unsupervised mining approaches of clustering can be incorporated in the segmentation of cells. In this study, we have developed and evaluated the performance of multiple unsupervised data mining techniques in cell image segmentation. We adapt four distinctive, yet complementary, methods for unsupervised learning, including those based on k-means clustering, EM, Otsu's threshold, and GMAC. Validation measures are defined, and the performance of the techniques is evaluated both quantitatively and qualitatively using synthetic and recently published real data. Experimental results demonstrate that k-means, Otsu's threshold, and GMAC perform similarly, and have more precise segmentation results than EM. We report that EM has higher recall values and lower precision results from under-segmentation due to its Gaussian model assumption. We also demonstrate that these methods need spatial information to segment complex real cell images with a high degree of efficacy, as expected in many medical informatics applications.

  17. The Hierarchical Spectral Merger Algorithm: A New Time Series Clustering Procedure

    KAUST Repository

    Euán, Carolina

    2018-04-12

    We present a new method for time series clustering which we call the Hierarchical Spectral Merger (HSM) method. This procedure is based on the spectral theory of time series and identifies series that share similar oscillations or waveforms. The extent of similarity between a pair of time series is measured using the total variation distance between their estimated spectral densities. At each step of the algorithm, every time two clusters merge, a new spectral density is estimated using the whole information present in both clusters, which is representative of all the series in the new cluster. The method is implemented in an R package HSMClust. We present two applications of the HSM method, one to data coming from wave-height measurements in oceanography and the other to electroencefalogram (EEG) data.

  18. Novel density-based and hierarchical density-based clustering algorithms for uncertain data.

    Science.gov (United States)

    Zhang, Xianchao; Liu, Han; Zhang, Xiaotong

    2017-09-01

    Uncertain data has posed a great challenge to traditional clustering algorithms. Recently, several algorithms have been proposed for clustering uncertain data, and among them density-based techniques seem promising for handling data uncertainty. However, some issues like losing uncertain information, high time complexity and nonadaptive threshold have not been addressed well in the previous density-based algorithm FDBSCAN and hierarchical density-based algorithm FOPTICS. In this paper, we firstly propose a novel density-based algorithm PDBSCAN, which improves the previous FDBSCAN from the following aspects: (1) it employs a more accurate method to compute the probability that the distance between two uncertain objects is less than or equal to a boundary value, instead of the sampling-based method in FDBSCAN; (2) it introduces new definitions of probability neighborhood, support degree, core object probability, direct reachability probability, thus reducing the complexity and solving the issue of nonadaptive threshold (for core object judgement) in FDBSCAN. Then, we modify the algorithm PDBSCAN to an improved version (PDBSCANi), by using a better cluster assignment strategy to ensure that every object will be assigned to the most appropriate cluster, thus solving the issue of nonadaptive threshold (for direct density reachability judgement) in FDBSCAN. Furthermore, as PDBSCAN and PDBSCANi have difficulties for clustering uncertain data with non-uniform cluster density, we propose a novel hierarchical density-based algorithm POPTICS by extending the definitions of PDBSCAN, adding new definitions of fuzzy core distance and fuzzy reachability distance, and employing a new clustering framework. POPTICS can reveal the cluster structures of the datasets with different local densities in different regions better than PDBSCAN and PDBSCANi, and it addresses the issues in FOPTICS. Experimental results demonstrate the superiority of our proposed algorithms over the existing

  19. GibbsCluster: unsupervised clustering and alignment of peptide sequences

    DEFF Research Database (Denmark)

    Andreatta, Massimo; Alvarez, Bruno; Nielsen, Morten

    2017-01-01

    motif characterizing each cluster. Several parameters are available to customize cluster analysis, including adjustable penalties for small clusters and overlapping groups and a trash cluster to remove outliers. As an example application, we used the server to deconvolute multiple specificities in large......-scale peptidome data generated by mass spectrometry. The server is available at http://www.cbs.dtu.dk/services/GibbsCluster-2.0....

  20. Unsupervised land cover change detection: meaningful sequential time series analysis

    CSIR Research Space (South Africa)

    Salmon, BP

    2011-06-01

    Full Text Available An automated land cover change detection method is proposed that uses coarse spatial resolution hyper-temporal earth observation satellite time series data. The study compared three different unsupervised clustering approaches that operate on short...

  1. Hierarchical Star Formation in Turbulent Media: Evidence from Young Star Clusters

    Energy Technology Data Exchange (ETDEWEB)

    Grasha, K.; Calzetti, D. [Astronomy Department, University of Massachusetts, Amherst, MA 01003 (United States); Elmegreen, B. G. [IBM Research Division, T.J. Watson Research Center, Yorktown Heights, NY (United States); Adamo, A.; Messa, M. [Department of Astronomy, The Oskar Klein Centre, Stockholm University, Stockholm (Sweden); Aloisi, A.; Bright, S. N.; Lee, J. C.; Ryon, J. E.; Ubeda, L. [Space Telescope Science Institute, Baltimore, MD (United States); Cook, D. O. [California Institute of Technology, 1200 East California Boulevard, Pasadena, CA (United States); Dale, D. A. [Department of Physics and Astronomy, University of Wyoming, Laramie, WY (United States); Fumagalli, M. [Institute for Computational Cosmology and Centre for Extragalactic Astronomy, Department of Physics, Durham University, Durham (United Kingdom); Gallagher III, J. S. [Department of Astronomy, University of Wisconsin–Madison, Madison, WI (United States); Gouliermis, D. A. [Zentrum für Astronomie der Universität Heidelberg, Institut für Theoretische Astrophysik, Albert-Ueberle-Str. 2, D-69120 Heidelberg (Germany); Grebel, E. K. [Astronomisches Rechen-Institut, Zentrum für Astronomie der Universität Heidelberg, Mönchhofstr. 12-14, D-69120, Heidelberg (Germany); Kahre, L. [Department of Astronomy, New Mexico State University, Las Cruces, NM (United States); Kim, H. [Gemini Observatory, La Serena (Chile); Krumholz, M. R., E-mail: kgrasha@astro.umass.edu [Research School of Astronomy and Astrophysics, Australian National University, Canberra, ACT 2611 (Australia)

    2017-06-10

    We present an analysis of the positions and ages of young star clusters in eight local galaxies to investigate the connection between the age difference and separation of cluster pairs. We find that star clusters do not form uniformly but instead are distributed so that the age difference increases with the cluster pair separation to the 0.25–0.6 power, and that the maximum size over which star formation is physically correlated ranges from ∼200 pc to ∼1 kpc. The observed trends between age difference and separation suggest that cluster formation is hierarchical both in space and time: clusters that are close to each other are more similar in age than clusters born further apart. The temporal correlations between stellar aggregates have slopes that are consistent with predictions of turbulence acting as the primary driver of star formation. The velocity associated with the maximum size is proportional to the galaxy’s shear, suggesting that the galactic environment influences the maximum size of the star-forming structures.

  2. Intensity-based hierarchical clustering in CT-scans: application to interactive segmentation in cardiology

    Science.gov (United States)

    Hadida, Jonathan; Desrosiers, Christian; Duong, Luc

    2011-03-01

    The segmentation of anatomical structures in Computed Tomography Angiography (CTA) is a pre-operative task useful in image guided surgery. Even though very robust and precise methods have been developed to help achieving a reliable segmentation (level sets, active contours, etc), it remains very time consuming both in terms of manual interactions and in terms of computation time. The goal of this study is to present a fast method to find coarse anatomical structures in CTA with few parameters, based on hierarchical clustering. The algorithm is organized as follows: first, a fast non-parametric histogram clustering method is proposed to compute a piecewise constant mask. A second step then indexes all the space-connected regions in the piecewise constant mask. Finally, a hierarchical clustering is achieved to build a graph representing the connections between the various regions in the piecewise constant mask. This step builds up a structural knowledge about the image. Several interactive features for segmentation are presented, for instance association or disassociation of anatomical structures. A comparison with the Mean-Shift algorithm is presented.

  3. A data-driven approach to estimating the number of clusters in hierarchical clustering [version 1; referees: 2 approved, 1 approved with reservations

    Directory of Open Access Journals (Sweden)

    Antoine E. Zambelli

    2016-12-01

    Full Text Available DNA microarray and gene expression problems often require a researcher to perform clustering on their data in a bid to better understand its structure. In cases where the number of clusters is not known, one can resort to hierarchical clustering methods. However, there currently exist very few automated algorithms for determining the true number of clusters in the data. We propose two new methods (mode and maximum difference for estimating the number of clusters in a hierarchical clustering framework to create a fully automated process with no human intervention. These methods are compared to the established elbow and gap statistic algorithms using simulated datasets and the Biobase Gene ExpressionSet. We also explore a data mixing procedure inspired by cross validation techniques. We find that the overall performance of the maximum difference method is comparable or greater to that of the gap statistic in multi-cluster scenarios, and achieves that performance at a fraction of the computational cost. This method also responds well to our mixing procedure, which opens the door to future research. We conclude that both the mode and maximum difference methods warrant further study related to their mixing and cross-validation potential. We particularly recommend the use of the maximum difference method in multi-cluster scenarios given its accuracy and execution times, and present it as an alternative to existing algorithms.

  4. GO-PCA: An Unsupervised Method to Explore Gene Expression Data Using Prior Knowledge.

    Science.gov (United States)

    Wagner, Florian

    2015-01-01

    Genome-wide expression profiling is a widely used approach for characterizing heterogeneous populations of cells, tissues, biopsies, or other biological specimen. The exploratory analysis of such data typically relies on generic unsupervised methods, e.g. principal component analysis (PCA) or hierarchical clustering. However, generic methods fail to exploit prior knowledge about the molecular functions of genes. Here, I introduce GO-PCA, an unsupervised method that combines PCA with nonparametric GO enrichment analysis, in order to systematically search for sets of genes that are both strongly correlated and closely functionally related. These gene sets are then used to automatically generate expression signatures with functional labels, which collectively aim to provide a readily interpretable representation of biologically relevant similarities and differences. The robustness of the results obtained can be assessed by bootstrapping. I first applied GO-PCA to datasets containing diverse hematopoietic cell types from human and mouse, respectively. In both cases, GO-PCA generated a small number of signatures that represented the majority of lineages present, and whose labels reflected their respective biological characteristics. I then applied GO-PCA to human glioblastoma (GBM) data, and recovered signatures associated with four out of five previously defined GBM subtypes. My results demonstrate that GO-PCA is a powerful and versatile exploratory method that reduces an expression matrix containing thousands of genes to a much smaller set of interpretable signatures. In this way, GO-PCA aims to facilitate hypothesis generation, design of further analyses, and functional comparisons across datasets.

  5. Prediction of in vitro and in vivo oestrogen receptor activity using hierarchical clustering

    Science.gov (United States)

    In this study, hierarchical clustering classification models were developed to predict in vitro and in vivo oestrogen receptor (ER) activity. Classification models were developed for binding, agonist, and antagonist in vitro ER activity and for mouse in vivo uterotrophic ER bindi...

  6. Kendall’s tau and agglomerative clustering for structure determination of hierarchical Archimedean copulas

    Directory of Open Access Journals (Sweden)

    Górecki J.

    2017-01-01

    Full Text Available Several successful approaches to structure determination of hierarchical Archimedean copulas (HACs proposed in the literature rely on agglomerative clustering and Kendall’s correlation coefficient. However, there has not been presented any theoretical proof justifying such approaches. This work fills this gap and introduces a theorem showing that, given the matrix of the pairwise Kendall correlation coefficients corresponding to a HAC, its structure can be recovered by an agglomerative clustering technique.

  7. Communication Base Station Log Analysis Based on Hierarchical Clustering

    Directory of Open Access Journals (Sweden)

    Zhang Shao-Hua

    2017-01-01

    Full Text Available Communication base stations generate massive data every day, these base station logs play an important value in mining of the business circles. This paper use data mining technology and hierarchical clustering algorithm to group the scope of business circle for the base station by recording the data of these base stations.Through analyzing the data of different business circle based on feature extraction and comparing different business circle category characteristics, which can choose a suitable area for operators of commercial marketing.

  8. Supervised versus unsupervised categorization: two sides of the same coin?

    Science.gov (United States)

    Pothos, Emmanuel M; Edwards, Darren J; Perlman, Amotz

    2011-09-01

    Supervised and unsupervised categorization have been studied in separate research traditions. A handful of studies have attempted to explore a possible convergence between the two. The present research builds on these studies, by comparing the unsupervised categorization results of Pothos et al. ( 2011 ; Pothos et al., 2008 ) with the results from two procedures of supervised categorization. In two experiments, we tested 375 participants with nine different stimulus sets and examined the relation between ease of learning of a classification, memory for a classification, and spontaneous preference for a classification. After taking into account the role of the number of category labels (clusters) in supervised learning, we found the three variables to be closely associated with each other. Our results provide encouragement for researchers seeking unified theoretical explanations for supervised and unsupervised categorization, but raise a range of challenging theoretical questions.

  9. Class imbalance in unsupervised change detection - A diagnostic analysis from urban remote sensing

    Science.gov (United States)

    Leichtle, Tobias; Geiß, Christian; Lakes, Tobia; Taubenböck, Hannes

    2017-08-01

    Automatic monitoring of changes on the Earth's surface is an intrinsic capability and simultaneously a persistent methodological challenge in remote sensing, especially regarding imagery with very-high spatial resolution (VHR) and complex urban environments. In order to enable a high level of automatization, the change detection problem is solved in an unsupervised way to alleviate efforts associated with collection of properly encoded prior knowledge. In this context, this paper systematically investigates the nature and effects of class distribution and class imbalance in an unsupervised binary change detection application based on VHR imagery over urban areas. For this purpose, a diagnostic framework for sensitivity analysis of a large range of possible degrees of class imbalance is presented, which is of particular importance with respect to unsupervised approaches where the content of images and thus the occurrence and the distribution of classes are generally unknown a priori. Furthermore, this framework can serve as a general technique to evaluate model transferability in any two-class classification problem. The applied change detection approach is based on object-based difference features calculated from VHR imagery and subsequent unsupervised two-class clustering using k-means, genetic k-means and self-organizing map (SOM) clustering. The results from two test sites with different structural characteristics of the built environment demonstrated that classification performance is generally worse in imbalanced class distribution settings while best results were reached in balanced or close to balanced situations. Regarding suitable accuracy measures for evaluating model performance in imbalanced settings, this study revealed that the Kappa statistics show significant response to class distribution while the true skill statistic was widely insensitive to imbalanced classes. In general, the genetic k-means clustering algorithm achieved the most robust results

  10. ClusterTAD: an unsupervised machine learning approach to detecting topologically associated domains of chromosomes from Hi-C data.

    Science.gov (United States)

    Oluwadare, Oluwatosin; Cheng, Jianlin

    2017-11-14

    With the development of chromosomal conformation capturing techniques, particularly, the Hi-C technique, the study of the spatial conformation of a genome is becoming an important topic in bioinformatics and computational biology. The Hi-C technique can generate genome-wide chromosomal interaction (contact) data, which can be used to investigate the higher-level organization of chromosomes, such as Topologically Associated Domains (TAD), i.e., locally packed chromosome regions bounded together by intra chromosomal contacts. The identification of the TADs for a genome is useful for studying gene regulation, genomic interaction, and genome function. Here, we formulate the TAD identification problem as an unsupervised machine learning (clustering) problem, and develop a new TAD identification method called ClusterTAD. We introduce a novel method to represent chromosomal contacts as features to be used by the clustering algorithm. Our results show that ClusterTAD can accurately predict the TADs on a simulated Hi-C data. Our method is also largely complementary and consistent with existing methods on the real Hi-C datasets of two mouse cells. The validation with the chromatin immunoprecipitation (ChIP) sequencing (ChIP-Seq) data shows that the domain boundaries identified by ClusterTAD have a high enrichment of CTCF binding sites, promoter-related marks, and enhancer-related histone modifications. As ClusterTAD is based on a proven clustering approach, it opens a new avenue to apply a large array of clustering methods developed in the machine learning field to the TAD identification problem. The source code, the results, and the TADs generated for the simulated and real Hi-C datasets are available here: https://github.com/BDM-Lab/ClusterTAD .

  11. Learning representation hierarchies by sharing visual features: a computational investigation of Persian character recognition with unsupervised deep learning.

    Science.gov (United States)

    Sadeghi, Zahra; Testolin, Alberto

    2017-08-01

    In humans, efficient recognition of written symbols is thought to rely on a hierarchical processing system, where simple features are progressively combined into more abstract, high-level representations. Here, we present a computational model of Persian character recognition based on deep belief networks, where increasingly more complex visual features emerge in a completely unsupervised manner by fitting a hierarchical generative model to the sensory data. Crucially, high-level internal representations emerging from unsupervised deep learning can be easily read out by a linear classifier, achieving state-of-the-art recognition accuracy. Furthermore, we tested the hypothesis that handwritten digits and letters share many common visual features: A generative model that captures the statistical structure of the letters distribution should therefore also support the recognition of written digits. To this aim, deep networks trained on Persian letters were used to build high-level representations of Persian digits, which were indeed read out with high accuracy. Our simulations show that complex visual features, such as those mediating the identification of Persian symbols, can emerge from unsupervised learning in multilayered neural networks and can support knowledge transfer across related domains.

  12. Unsupervised Approach Data Analysis Based on Fuzzy Possibilistic Clustering: Application to Medical Image MRI

    Directory of Open Access Journals (Sweden)

    Nour-Eddine El Harchaoui

    2013-01-01

    Full Text Available The analysis and processing of large data are a challenge for researchers. Several approaches have been used to model these complex data, and they are based on some mathematical theories: fuzzy, probabilistic, possibilistic, and evidence theories. In this work, we propose a new unsupervised classification approach that combines the fuzzy and possibilistic theories; our purpose is to overcome the problems of uncertain data in complex systems. We used the membership function of fuzzy c-means (FCM to initialize the parameters of possibilistic c-means (PCM, in order to solve the problem of coinciding clusters that are generated by PCM and also overcome the weakness of FCM to noise. To validate our approach, we used several validity indexes and we compared them with other conventional classification algorithms: fuzzy c-means, possibilistic c-means, and possibilistic fuzzy c-means. The experiments were realized on different synthetics data sets and real brain MR images.

  13. Application of unsupervised pattern recognition approaches for exploration of rare earth elements in Se-Chahun iron ore, central Iran

    Science.gov (United States)

    Sarparandeh, Mohammadali; Hezarkhani, Ardeshir

    2017-12-01

    The use of efficient methods for data processing has always been of interest to researchers in the field of earth sciences. Pattern recognition techniques are appropriate methods for high-dimensional data such as geochemical data. Evaluation of the geochemical distribution of rare earth elements (REEs) requires the use of such methods. In particular, the multivariate nature of REE data makes them a good target for numerical analysis. The main subject of this paper is application of unsupervised pattern recognition approaches in evaluating geochemical distribution of REEs in the Kiruna type magnetite-apatite deposit of Se-Chahun. For this purpose, 42 bulk lithology samples were collected from the Se-Chahun iron ore deposit. In this study, 14 rare earth elements were measured with inductively coupled plasma mass spectrometry (ICP-MS). Pattern recognition makes it possible to evaluate the relations between the samples based on all these 14 features, simultaneously. In addition to providing easy solutions, discovery of the hidden information and relations of data samples is the advantage of these methods. Therefore, four clustering methods (unsupervised pattern recognition) - including a modified basic sequential algorithmic scheme (MBSAS), hierarchical (agglomerative) clustering, k-means clustering and self-organizing map (SOM) - were applied and results were evaluated using the silhouette criterion. Samples were clustered in four types. Finally, the results of this study were validated with geological facts and analysis results from, for example, scanning electron microscopy (SEM), X-ray diffraction (XRD), ICP-MS and optical mineralogy. The results of the k-means clustering and SOM methods have the best matches with reality, with experimental studies of samples and with field surveys. Since only the rare earth elements are used in this division, a good agreement of the results with lithology is considerable. It is concluded that the combination of the proposed

  14. Radar Emission Sources Identification Based on Hierarchical Agglomerative Clustering for Large Data Sets

    Directory of Open Access Journals (Sweden)

    Janusz Dudczyk

    2016-01-01

    Full Text Available More advanced recognition methods, which may recognize particular copies of radars of the same type, are called identification. The identification process of radar devices is a more specialized task which requires methods based on the analysis of distinctive features. These features are distinguished from the signals coming from the identified devices. Such a process is called Specific Emitter Identification (SEI. The identification of radar emission sources with the use of classic techniques based on the statistical analysis of basic measurable parameters of a signal such as Radio Frequency, Amplitude, Pulse Width, or Pulse Repetition Interval is not sufficient for SEI problems. This paper presents the method of hierarchical data clustering which is used in the process of radar identification. The Hierarchical Agglomerative Clustering Algorithm (HACA based on Generalized Agglomerative Scheme (GAS implemented and used in the research method is parameterized; therefore, it is possible to compare the results. The results of clustering are presented in dendrograms in this paper. The received results of grouping and identification based on HACA are compared with other SEI methods in order to assess the degree of their usefulness and effectiveness for systems of ESM/ELINT class.

  15. Modeling Language and Cognition with Deep Unsupervised Learning:A Tutorial Overview

    OpenAIRE

    Marco eZorzi; Marco eZorzi; Alberto eTestolin; Ivilin Peev Stoianov; Ivilin Peev Stoianov

    2013-01-01

    Deep unsupervised learning in stochastic recurrent neural networks with many layers of hidden units is a recent breakthrough in neural computation research. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. In this article we discuss the theoretical foundations of this approach and we review key issues related to training, testing and analysis of deep networks for modeling language and cog...

  16. Modeling language and cognition with deep unsupervised learning: a tutorial overview

    OpenAIRE

    Zorzi, Marco; Testolin, Alberto; Stoianov, Ivilin P.

    2013-01-01

    Deep unsupervised learning in stochastic recurrent neural networks with many layers of hidden units is a recent breakthrough in neural computation research. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. In this article we discuss the theoretical foundations of this approach and we review key issues related to training, testing and analysis of deep networks for modeling language and cog...

  17. Evolutionary-Hierarchical Bases of the Formation of Cluster Model of Innovation Economic Development

    Directory of Open Access Journals (Sweden)

    Yuliya Vladimirovna Dubrovskaya

    2016-10-01

    Full Text Available The functioning of a modern economic system is based on the interaction of objects of different hierarchical levels. Thus, the problem of the study of innovation processes taking into account the mutual influence of the activities of these economic actors becomes important. The paper dwells evolutionary basis for the formation of models of innovation development on the basis of micro and macroeconomic analysis. Most of the concepts recognized that despite a big number of diverse models, the coordination of the relations between economic agents is of crucial importance for the successful innovation development. According to the results of the evolutionary-hierarchical analysis, the authors reveal key phases of the development of forms of business cooperation, science and government in the domestic economy. It has become the starting point of the conception of the characteristics of the interaction in the cluster models of innovation development of the economy. Considerable expectancies on improvement of the national innovative system are connected with the development of cluster and network structures. The main objective of government authorities is the formation of mechanisms and institutions that will foster cooperation between members of the clusters. The article explains that the clusters cannot become the factors in the growth of the national economy, not being an effective tool for interaction between the actors of the regional innovative systems.

  18. Efficient algorithms for accurate hierarchical clustering of huge datasets: tackling the entire protein space

    OpenAIRE

    Loewenstein, Yaniv; Portugaly, Elon; Fromer, Menachem; Linial, Michal

    2008-01-01

    Motivation: UPGMA (average linking) is probably the most popular algorithm for hierarchical data clustering, especially in computational biology. However, UPGMA requires the entire dissimilarity matrix in memory. Due to this prohibitive requirement, UPGMA is not scalable to very large datasets. Application: We present a novel class of memory-constrained UPGMA (MC-UPGMA) algorithms. Given any practical memory size constraint, this framework guarantees the correct clustering solution without ex...

  19. Towards Statistical Unsupervised Online Learning for Music Listening with Hearing Devices

    DEFF Research Database (Denmark)

    Purwins, Hendrik; Marchini, Marco; Marxer, Richard

    of sounds into phonetic/instrument categories and learning of instrument event sequences is performed jointly using a Hierarchical Dirichlet Process Hidden Markov Model. Whereas machines often learn by processing a large data base and subsequently updating parameters of the algorithm, humans learn...... and their respective transition counts. We propose to use online learning for the co-evolution of both CI user and machine in (re-)learning musical language. [1] Marco Marchini and Hendrik Purwins. Unsupervised analysis and generation of audio percussion sequences. In International Symposium on Computer Music Modeling...... categories) as well as the temporal context horizon (e.g. storing up to 2-note sequences or up to 10-note sequences) is adaptable. The framework in [1] is based on two cognitively plausible principles: unsupervised learning and statistical learning. Opposed to supervised learning in primary school children...

  20. Water quality assessment with hierarchical cluster analysis based on Mahalanobis distance.

    Science.gov (United States)

    Du, Xiangjun; Shao, Fengjing; Wu, Shunyao; Zhang, Hanlin; Xu, Si

    2017-07-01

    Water quality assessment is crucial for assessment of marine eutrophication, prediction of harmful algal blooms, and environment protection. Previous studies have developed many numeric modeling methods and data driven approaches for water quality assessment. The cluster analysis, an approach widely used for grouping data, has also been employed. However, there are complex correlations between water quality variables, which play important roles in water quality assessment but have always been overlooked. In this paper, we analyze correlations between water quality variables and propose an alternative method for water quality assessment with hierarchical cluster analysis based on Mahalanobis distance. Further, we cluster water quality data collected form coastal water of Bohai Sea and North Yellow Sea of China, and apply clustering results to evaluate its water quality. To evaluate the validity, we also cluster the water quality data with cluster analysis based on Euclidean distance, which are widely adopted by previous studies. The results show that our method is more suitable for water quality assessment with many correlated water quality variables. To our knowledge, it is the first attempt to apply Mahalanobis distance for coastal water quality assessment.

  1. Canonical PSO Based K-Means Clustering Approach for Real Datasets.

    Science.gov (United States)

    Dey, Lopamudra; Chakraborty, Sanjay

    2014-01-01

    "Clustering" the significance and application of this technique is spread over various fields. Clustering is an unsupervised process in data mining, that is why the proper evaluation of the results and measuring the compactness and separability of the clusters are important issues. The procedure of evaluating the results of a clustering algorithm is known as cluster validity measure. Different types of indexes are used to solve different types of problems and indices selection depends on the kind of available data. This paper first proposes Canonical PSO based K-means clustering algorithm and also analyses some important clustering indices (intercluster, intracluster) and then evaluates the effects of those indices on real-time air pollution database, wholesale customer, wine, and vehicle datasets using typical K-means, Canonical PSO based K-means, simple PSO based K-means, DBSCAN, and Hierarchical clustering algorithms. This paper also describes the nature of the clusters and finally compares the performances of these clustering algorithms according to the validity assessment. It also defines which algorithm will be more desirable among all these algorithms to make proper compact clusters on this particular real life datasets. It actually deals with the behaviour of these clustering algorithms with respect to validation indexes and represents their results of evaluation in terms of mathematical and graphical forms.

  2. Multi-documents summarization based on clustering of learning object using hierarchical clustering

    Science.gov (United States)

    Mustamiin, M.; Budi, I.; Santoso, H. B.

    2018-03-01

    The Open Educational Resources (OER) is a portal of teaching, learning and research resources that is available in public domain and freely accessible. Learning contents or Learning Objects (LO) are granular and can be reused for constructing new learning materials. LO ontology-based searching techniques can be used to search for LO in the Indonesia OER. In this research, LO from search results are used as an ingredient to create new learning materials according to the topic searched by users. Summarizing-based grouping of LO use Hierarchical Agglomerative Clustering (HAC) with the dependency context to the user’s query which has an average value F-Measure of 0.487, while summarizing by K-Means F-Measure only has an average value of 0.336.

  3. An Energy Efficient Cooperative Hierarchical MIMO Clustering Scheme for Wireless Sensor Networks

    Directory of Open Access Journals (Sweden)

    Sungyoung Lee

    2011-12-01

    Full Text Available In this work, we present an energy efficient hierarchical cooperative clustering scheme for wireless sensor networks. Communication cost is a crucial factor in depleting the energy of sensor nodes. In the proposed scheme, nodes cooperate to form clusters at each level of network hierarchy ensuring maximal coverage and minimal energy expenditure with relatively uniform distribution of load within the network. Performance is enhanced by cooperative multiple-input multiple-output (MIMO communication ensuring energy efficiency for WSN deployments over large geographical areas. We test our scheme using TOSSIM and compare the proposed scheme with cooperative multiple-input multiple-output (CMIMO clustering scheme and traditional multihop Single-Input-Single-Output (SISO routing approach. Performance is evaluated on the basis of number of clusters, number of hops, energy consumption and network lifetime. Experimental results show significant energy conservation and increase in network lifetime as compared to existing schemes.

  4. Hierarchical clustering analysis of blood plasma lipidomics profiles from mono- and dizygotic twin families

    NARCIS (Netherlands)

    Draisma, H.H.; Reijmers, T.H.; Meulman, J.J.; Greef, J. van der; Hankemeier, T.; Boomsma, D.I.

    2013-01-01

    Twin and family studies are typically used to elucidate the relative contribution of genetic and environmental variation to phenotypic variation. Here, we apply a quantitative genetic method based on hierarchical clustering, to blood plasma lipidomics data obtained in a healthy cohort consisting of

  5. On the Disruption of Star Clusters in a Hierarchical Interstellar Medium

    Science.gov (United States)

    Elmegreen, Bruce G.; Hunter, Deidre A.

    2010-03-01

    The distribution of the number of clusters as a function of mass M and age T suggests that clusters get eroded or dispersed in a regular way over time, such that the cluster number decreases inversely as an approximate power law with T within each fixed interval of M. This power law is inconsistent with standard dispersal mechanisms such as cluster evaporation and cloud collisions. In the conventional interpretation, it requires the unlikely situation where diverse mechanisms stitch together over time in a way that is independent of environment or M. Here, we consider another model in which the large-scale distribution of gas in each star-forming region plays an important role. We note that star clusters form with positional and temporal correlations in giant cloud complexes, and suggest that these complexes dominate the tidal force and collisional influence on a cluster during its first several hundred million years. Because the cloud complex density decreases regularly with position from the cluster birth site, the harassment and collision rates between the cluster and the cloud pieces decrease regularly with age as the cluster drifts. This decrease is typically a power law of the form required to explain the mass-age distribution. We reproduce this distribution for a variety of cases, including rapid disruption, slow erosion, combinations of these two, cluster-cloud collisions, cluster disruption by hierarchical disassembly, and partial cluster disruption. We also consider apparent cluster mass loss by fading below the surface brightness limit of a survey. In all cases, the observed log M-log T diagram can be reproduced under reasonable assumptions.

  6. ON THE DISRUPTION OF STAR CLUSTERS IN A HIERARCHICAL INTERSTELLAR MEDIUM

    International Nuclear Information System (INIS)

    Elmegreen, Bruce G.; Hunter, Deidre A.

    2010-01-01

    The distribution of the number of clusters as a function of mass M and age T suggests that clusters get eroded or dispersed in a regular way over time, such that the cluster number decreases inversely as an approximate power law with T within each fixed interval of M. This power law is inconsistent with standard dispersal mechanisms such as cluster evaporation and cloud collisions. In the conventional interpretation, it requires the unlikely situation where diverse mechanisms stitch together over time in a way that is independent of environment or M. Here, we consider another model in which the large-scale distribution of gas in each star-forming region plays an important role. We note that star clusters form with positional and temporal correlations in giant cloud complexes, and suggest that these complexes dominate the tidal force and collisional influence on a cluster during its first several hundred million years. Because the cloud complex density decreases regularly with position from the cluster birth site, the harassment and collision rates between the cluster and the cloud pieces decrease regularly with age as the cluster drifts. This decrease is typically a power law of the form required to explain the mass-age distribution. We reproduce this distribution for a variety of cases, including rapid disruption, slow erosion, combinations of these two, cluster-cloud collisions, cluster disruption by hierarchical disassembly, and partial cluster disruption. We also consider apparent cluster mass loss by fading below the surface brightness limit of a survey. In all cases, the observed log M-log T diagram can be reproduced under reasonable assumptions.

  7. The Hierarchical Clustering of Tax Burden in the EU27

    Directory of Open Access Journals (Sweden)

    Simkova Nikola

    2015-09-01

    Full Text Available The issue of taxation has become more important due to a significant share of the government revenue. There are several ways of expressing the tax burden of countries. This paper describes the traditional approach as a share of tax revenue to GDP which is applied to the total taxation and the capital taxation as a part of tax systems affecting investment decisions. The implicit tax rate on capital created by Eurostat also offers a possible explanation of the tax burden on capital, so its components are analysed in detail. This study uses one of the econometric methods called the hierarchical clustering. The data on which the clustering is based comprises countries in the EU27 for the period of 1995 – 2012. The aim of this paper is to reveal clusters of countries in the EU27 with similar tax burden or tax changes. The findings suggest that mainly newly acceding countries (2004 and 2007 are in a group of countries with a low tax burden which tried to encourage investors by favourable tax rates. On the other hand, there are mostly countries from the original EU15. Some clusters may be explained by similar historical development, geographic and demographic characteristics.

  8. Clustering and visualizing similarity networks of membrane proteins.

    Science.gov (United States)

    Hu, Geng-Ming; Mai, Te-Lun; Chen, Chi-Ming

    2015-08-01

    We proposed a fast and unsupervised clustering method, minimum span clustering (MSC), for analyzing the sequence-structure-function relationship of biological networks, and demonstrated its validity in clustering the sequence/structure similarity networks (SSN) of 682 membrane protein (MP) chains. The MSC clustering of MPs based on their sequence information was found to be consistent with their tertiary structures and functions. For the largest seven clusters predicted by MSC, the consistency in chain function within the same cluster is found to be 100%. From analyzing the edge distribution of SSN for MPs, we found a characteristic threshold distance for the boundary between clusters, over which SSN of MPs could be properly clustered by an unsupervised sparsification of the network distance matrix. The clustering results of MPs from both MSC and the unsupervised sparsification methods are consistent with each other, and have high intracluster similarity and low intercluster similarity in sequence, structure, and function. Our study showed a strong sequence-structure-function relationship of MPs. We discussed evidence of convergent evolution of MPs and suggested applications in finding structural similarities and predicting biological functions of MP chains based on their sequence information. © 2015 Wiley Periodicals, Inc.

  9. ESPRIT-Tree: hierarchical clustering analysis of millions of 16S rRNA pyrosequences in quasilinear computational time.

    Science.gov (United States)

    Cai, Yunpeng; Sun, Yijun

    2011-08-01

    Taxonomy-independent analysis plays an essential role in microbial community analysis. Hierarchical clustering is one of the most widely employed approaches to finding operational taxonomic units, the basis for many downstream analyses. Most existing algorithms have quadratic space and computational complexities, and thus can be used only for small or medium-scale problems. We propose a new online learning-based algorithm that simultaneously addresses the space and computational issues of prior work. The basic idea is to partition a sequence space into a set of subspaces using a partition tree constructed using a pseudometric, then recursively refine a clustering structure in these subspaces. The technique relies on new methods for fast closest-pair searching and efficient dynamic insertion and deletion of tree nodes. To avoid exhaustive computation of pairwise distances between clusters, we represent each cluster of sequences as a probabilistic sequence, and define a set of operations to align these probabilistic sequences and compute genetic distances between them. We present analyses of space and computational complexity, and demonstrate the effectiveness of our new algorithm using a human gut microbiota data set with over one million sequences. The new algorithm exhibits a quasilinear time and space complexity comparable to greedy heuristic clustering algorithms, while achieving a similar accuracy to the standard hierarchical clustering algorithm.

  10. Hierarchical Clustering of Large Databases and Classification of Antibiotics at High Noise Levels

    Directory of Open Access Journals (Sweden)

    Alexander V. Yarkov

    2008-12-01

    Full Text Available A new algorithm for divisive hierarchical clustering of chemical compounds based on 2D structural fragments is suggested. The algorithm is deterministic, and given a random ordering of the input, will always give the same clustering and can process a database up to 2 million records on a standard PC. The algorithm was used for classification of 1,183 antibiotics mixed with 999,994 random chemical structures. Similarity threshold, at which best separation of active and non active compounds took place, was estimated as 0.6. 85.7% of the antibiotics were successfully classified at this threshold with 0.4% of inaccurate compounds. A .sdf file was created with the probe molecules for clustering of external databases.

  11. Unsupervised learning of mixture models based on swarm intelligence and neural networks with optimal completion using incomplete data

    Directory of Open Access Journals (Sweden)

    Ahmed R. Abas

    2012-07-01

    Full Text Available In this paper, a new algorithm is presented for unsupervised learning of finite mixture models (FMMs using data set with missing values. This algorithm overcomes the local optima problem of the Expectation-Maximization (EM algorithm via integrating the EM algorithm with Particle Swarm Optimization (PSO. In addition, the proposed algorithm overcomes the problem of biased estimation due to overlapping clusters in estimating missing values in the input data set by integrating locally-tuned general regression neural networks with Optimal Completion Strategy (OCS. A comparison study shows the superiority of the proposed algorithm over other algorithms commonly used in the literature in unsupervised learning of FMM parameters that result in minimum mis-classification errors when used in clustering incomplete data set that is generated from overlapping clusters and these clusters are largely different in their sizes.

  12. Hierarchical clustering into groups of human brain regions according to elemental composition

    International Nuclear Information System (INIS)

    Stedman, J.D.; Spyrou, N.M.

    1998-01-01

    Thirteen brain regions were dissected from both hemispheres of fifteen 'normal' ageing subjects (8 females, 7 males) of mean age 79±7 years. Elemental compositions were determined by simultaneous application of particle induced X-ray emission (PIXE) and Rutherford backscattering (RBS) analyses using a 2 MeV, 4 nA proton beam scanned over 4 mm 2 of the sample surface. Elemental concentrations were found to be dependent upon the brain region and hemisphere studied. Hierarchical cluster analysis was applied to group the brain regions according to the sample concentrations of eight elements. The resulting dendrogram is presented and its clusters related to the sample compositions of grey and white matter. (author)

  13. Hierarchical Bayesian modelling of gene expression time series across irregularly sampled replicates and clusters.

    Science.gov (United States)

    Hensman, James; Lawrence, Neil D; Rattray, Magnus

    2013-08-20

    Time course data from microarrays and high-throughput sequencing experiments require simple, computationally efficient and powerful statistical models to extract meaningful biological signal, and for tasks such as data fusion and clustering. Existing methodologies fail to capture either the temporal or replicated nature of the experiments, and often impose constraints on the data collection process, such as regularly spaced samples, or similar sampling schema across replications. We propose hierarchical Gaussian processes as a general model of gene expression time-series, with application to a variety of problems. In particular, we illustrate the method's capacity for missing data imputation, data fusion and clustering.The method can impute data which is missing both systematically and at random: in a hold-out test on real data, performance is significantly better than commonly used imputation methods. The method's ability to model inter- and intra-cluster variance leads to more biologically meaningful clusters. The approach removes the necessity for evenly spaced samples, an advantage illustrated on a developmental Drosophila dataset with irregular replications. The hierarchical Gaussian process model provides an excellent statistical basis for several gene-expression time-series tasks. It has only a few additional parameters over a regular GP, has negligible additional complexity, is easily implemented and can be integrated into several existing algorithms. Our experiments were implemented in python, and are available from the authors' website: http://staffwww.dcs.shef.ac.uk/people/J.Hensman/.

  14. A new clustering algorithm for scanning electron microscope images

    Science.gov (United States)

    Yousef, Amr; Duraisamy, Prakash; Karim, Mohammad

    2016-04-01

    A scanning electron microscope (SEM) is a type of electron microscope that produces images of a sample by scanning it with a focused beam of electrons. The electrons interact with the sample atoms, producing various signals that are collected by detectors. The gathered signals contain information about the sample's surface topography and composition. The electron beam is generally scanned in a raster scan pattern, and the beam's position is combined with the detected signal to produce an image. The most common configuration for an SEM produces a single value per pixel, with the results usually rendered as grayscale images. The captured images may be produced with insufficient brightness, anomalous contrast, jagged edges, and poor quality due to low signal-to-noise ratio, grained topography and poor surface details. The segmentation of the SEM images is a tackling problems in the presence of the previously mentioned distortions. In this paper, we are stressing on the clustering of these type of images. In that sense, we evaluate the performance of the well-known unsupervised clustering and classification techniques such as connectivity based clustering (hierarchical clustering), centroid-based clustering, distribution-based clustering and density-based clustering. Furthermore, we propose a new spatial fuzzy clustering technique that works efficiently on this type of images and compare its results against these regular techniques in terms of clustering validation metrics.

  15. Rough-fuzzy clustering and unsupervised feature selection for wavelet based MR image segmentation.

    Directory of Open Access Journals (Sweden)

    Pradipta Maji

    Full Text Available Image segmentation is an indispensable process in the visualization of human tissues, particularly during clinical analysis of brain magnetic resonance (MR images. For many human experts, manual segmentation is a difficult and time consuming task, which makes an automated brain MR image segmentation method desirable. In this regard, this paper presents a new segmentation method for brain MR images, integrating judiciously the merits of rough-fuzzy computing and multiresolution image analysis technique. The proposed method assumes that the major brain tissues, namely, gray matter, white matter, and cerebrospinal fluid from the MR images are considered to have different textural properties. The dyadic wavelet analysis is used to extract the scale-space feature vector for each pixel, while the rough-fuzzy clustering is used to address the uncertainty problem of brain MR image segmentation. An unsupervised feature selection method is introduced, based on maximum relevance-maximum significance criterion, to select relevant and significant textural features for segmentation problem, while the mathematical morphology based skull stripping preprocessing step is proposed to remove the non-cerebral tissues like skull. The performance of the proposed method, along with a comparison with related approaches, is demonstrated on a set of synthetic and real brain MR images using standard validity indices.

  16. CHIMERA: Top-down model for hierarchical, overlapping and directed cluster structures in directed and weighted complex networks

    Science.gov (United States)

    Franke, R.

    2016-11-01

    In many networks discovered in biology, medicine, neuroscience and other disciplines special properties like a certain degree distribution and hierarchical cluster structure (also called communities) can be observed as general organizing principles. Detecting the cluster structure of an unknown network promises to identify functional subdivisions, hierarchy and interactions on a mesoscale. It is not trivial choosing an appropriate detection algorithm because there are multiple network, cluster and algorithmic properties to be considered. Edges can be weighted and/or directed, clusters overlap or build a hierarchy in several ways. Algorithms differ not only in runtime, memory requirements but also in allowed network and cluster properties. They are based on a specific definition of what a cluster is, too. On the one hand, a comprehensive network creation model is needed to build a large variety of benchmark networks with different reasonable structures to compare algorithms. On the other hand, if a cluster structure is already known, it is desirable to separate effects of this structure from other network properties. This can be done with null model networks that mimic an observed cluster structure to improve statistics on other network features. A third important application is the general study of properties in networks with different cluster structures, possibly evolving over time. Currently there are good benchmark and creation models available. But what is left is a precise sandbox model to build hierarchical, overlapping and directed clusters for undirected or directed, binary or weighted complex random networks on basis of a sophisticated blueprint. This gap shall be closed by the model CHIMERA (Cluster Hierarchy Interconnection Model for Evaluation, Research and Analysis) which will be introduced and described here for the first time.

  17. Integrating the Supervised Information into Unsupervised Learning

    Directory of Open Access Journals (Sweden)

    Ping Ling

    2013-01-01

    Full Text Available This paper presents an assembling unsupervised learning framework that adopts the information coming from the supervised learning process and gives the corresponding implementation algorithm. The algorithm consists of two phases: extracting and clustering data representatives (DRs firstly to obtain labeled training data and then classifying non-DRs based on labeled DRs. The implementation algorithm is called SDSN since it employs the tuning-scaled Support vector domain description to collect DRs, uses spectrum-based method to cluster DRs, and adopts the nearest neighbor classifier to label non-DRs. The validation of the clustering procedure of the first-phase is analyzed theoretically. A new metric is defined data dependently in the second phase to allow the nearest neighbor classifier to work with the informed information. A fast training approach for DRs’ extraction is provided to bring more efficiency. Experimental results on synthetic and real datasets verify that the proposed idea is of correctness and performance and SDSN exhibits higher popularity in practice over the traditional pure clustering procedure.

  18. Hierarchical and Non-Hierarchical Linear and Non-Linear Clustering Methods to “Shakespeare Authorship Question”

    Directory of Open Access Journals (Sweden)

    Refat Aljumily

    2015-09-01

    Full Text Available A few literary scholars have long claimed that Shakespeare did not write some of his best plays (history plays and tragedies and proposed at one time or another various suspect authorship candidates. Most modern-day scholars of Shakespeare have rejected this claim, arguing that strong evidence that Shakespeare wrote the plays and poems being his name appears on them as the author. This has caused and led to an ongoing scholarly academic debate for quite some long time. Stylometry is a fast-growing field often used to attribute authorship to anonymous or disputed texts. Stylometric attempts to resolve this literary puzzle have raised interesting questions over the past few years. The following paper contributes to “the Shakespeare authorship question” by using a mathematically-based methodology to examine the hypothesis that Shakespeare wrote all the disputed plays traditionally attributed to him. More specifically, the mathematically based methodology used here is based on Mean Proximity, as a linear hierarchical clustering method, and on Principal Components Analysis, as a non-hierarchical linear clustering method. It is also based, for the first time in the domain, on Self-Organizing Map U-Matrix and Voronoi Map, as non-linear clustering methods to cover the possibility that our data contains significant non-linearities. Vector Space Model (VSM is used to convert texts into vectors in a high dimensional space. The aim of which is to compare the degrees of similarity within and between limited samples of text (the disputed plays. The various works and plays assumed to have been written by Shakespeare and possible authors notably, Sir Francis Bacon, Christopher Marlowe, John Fletcher, and Thomas Kyd, where “similarity” is defined in terms of correlation/distance coefficient measure based on the frequency of usage profiles of function words, word bi-grams, and character triple-grams. The claim that Shakespeare authored all the disputed

  19. Classification Of Cluster Area Forsatellite Image

    Directory of Open Access Journals (Sweden)

    Thwe Zin Phyo

    2015-06-01

    Full Text Available Abstract This paper describes area classification for Landsat7 satellite image. The main purpose of this system is to classify the area of each cluster contained in a satellite image. To classify this image firstly need to clusterthe satellite image into different land cover types. Clustering is an unsupervised learning method that aimsto classify an image into homogeneous regions. This system is implemented based on color features with K-means clustering unsupervised algorithm. This method does not need to train image before clustering.The clusters of satellite image are grouped into a set of three clusters for Landsat7 satellite image. For this work the combined band 432 from Landsat7 satellite is used as an input. Satellite imageMandalay area in 2001 is chosen to test the segmentation method. After clustering a specific range for three clustered images must be defined in order to obtain greenland water and urbanbalance.This system is implemented by using MATLAB programming language.

  20. Assessment of genetic divergence in tomato through agglomerative hierarchical clustering and principal component analysis

    International Nuclear Information System (INIS)

    Iqbal, Q.; Saleem, M.Y.; Hameed, A.; Asghar, M.

    2014-01-01

    For the improvement of qualitative and quantitative traits, existence of variability has prime importance in plant breeding. Data on different morphological and reproductive traits of 47 tomato genotypes were analyzed for correlation,agglomerative hierarchical clustering and principal component analysis (PCA) to select genotypes and traits for future breeding program. Correlation analysis revealed significant positive association between yield and yield components like fruit diameter, single fruit weight and number of fruits plant-1. Principal component (PC) analysis depicted first three PCs with Eigen-value higher than 1 contributing 81.72% of total variability for different traits. The PC-I showed positive factor loadings for all the traits except number of fruits plant-1. The contribution of single fruit weight and fruit diameter was highest in PC-1. Cluster analysis grouped all genotypes into five divergent clusters. The genotypes in cluster-II and cluster-V exhibited uniform maturity and higher yield. The D2 statistics confirmed highest distance between cluster- III and cluster-V while maximum similarity was observed in cluster-II and cluster-III. It is therefore suggested that crosses between genotypes of cluster-II and cluster-V with those of cluster-I and cluster-III may exhibit heterosis in F1 for hybrid breeding and for selection of superior genotypes in succeeding generations for cross breeding programme. (author)

  1. Evaluation of hierarchical agglomerative cluster analysis methods for discrimination of primary biological aerosol

    Directory of Open Access Journals (Sweden)

    I. Crawford

    2015-11-01

    Full Text Available In this paper we present improved methods for discriminating and quantifying primary biological aerosol particles (PBAPs by applying hierarchical agglomerative cluster analysis to multi-parameter ultraviolet-light-induced fluorescence (UV-LIF spectrometer data. The methods employed in this study can be applied to data sets in excess of 1 × 106 points on a desktop computer, allowing for each fluorescent particle in a data set to be explicitly clustered. This reduces the potential for misattribution found in subsampling and comparative attribution methods used in previous approaches, improving our capacity to discriminate and quantify PBAP meta-classes. We evaluate the performance of several hierarchical agglomerative cluster analysis linkages and data normalisation methods using laboratory samples of known particle types and an ambient data set. Fluorescent and non-fluorescent polystyrene latex spheres were sampled with a Wideband Integrated Bioaerosol Spectrometer (WIBS-4 where the optical size, asymmetry factor and fluorescent measurements were used as inputs to the analysis package. It was found that the Ward linkage with z-score or range normalisation performed best, correctly attributing 98 and 98.1 % of the data points respectively. The best-performing methods were applied to the BEACHON-RoMBAS (Bio–hydro–atmosphere interactions of Energy, Aerosols, Carbon, H2O, Organics and Nitrogen–Rocky Mountain Biogenic Aerosol Study ambient data set, where it was found that the z-score and range normalisation methods yield similar results, with each method producing clusters representative of fungal spores and bacterial aerosol, consistent with previous results. The z-score result was compared to clusters generated with previous approaches (WIBS AnalysiS Program, WASP where we observe that the subsampling and comparative attribution method employed by WASP results in the overestimation of the fungal spore concentration by a factor of 1.5 and the

  2. The use of hierarchical clustering for the design of optimized monitoring networks

    Science.gov (United States)

    Soares, Joana; Makar, Paul Andrew; Aklilu, Yayne; Akingunola, Ayodeji

    2018-05-01

    Associativity analysis is a powerful tool to deal with large-scale datasets by clustering the data on the basis of (dis)similarity and can be used to assess the efficacy and design of air quality monitoring networks. We describe here our use of Kolmogorov-Zurbenko filtering and hierarchical clustering of NO2 and SO2 passive and continuous monitoring data to analyse and optimize air quality networks for these species in the province of Alberta, Canada. The methodology applied in this study assesses dissimilarity between monitoring station time series based on two metrics: 1 - R, R being the Pearson correlation coefficient, and the Euclidean distance; we find that both should be used in evaluating monitoring site similarity. We have combined the analytic power of hierarchical clustering with the spatial information provided by deterministic air quality model results, using the gridded time series of model output as potential station locations, as a proxy for assessing monitoring network design and for network optimization. We demonstrate that clustering results depend on the air contaminant analysed, reflecting the difference in the respective emission sources of SO2 and NO2 in the region under study. Our work shows that much of the signal identifying the sources of NO2 and SO2 emissions resides in shorter timescales (hourly to daily) due to short-term variation of concentrations and that longer-term averages in data collection may lose the information needed to identify local sources. However, the methodology identifies stations mainly influenced by seasonality, if larger timescales (weekly to monthly) are considered. We have performed the first dissimilarity analysis based on gridded air quality model output and have shown that the methodology is capable of generating maps of subregions within which a single station will represent the entire subregion, to a given level of dissimilarity. We have also shown that our approach is capable of identifying different

  3. An Algorithm for Inspecting Self Check-in Airline Luggage Based on Hierarchical Clustering and Cube-fitting

    Directory of Open Access Journals (Sweden)

    Gao Qingji

    2014-04-01

    Full Text Available Airport passengers are required to put only one baggage each time in the check-in self-service so that the baggage can be detected and identified successfully. In order to automatically get the number of baggage that had been put on the conveyor belt, dual laser rangefinders are used to scan the outer contour of luggage in this paper. The algorithm based on hierarchical clustering and cube-fitting is proposed to inspect the number and dimension of airline luggage. Firstly, the point cloud is projected to vertical direction. By the analysis of one-dimensional clustering, the number and height of luggage will be quickly computed. Secondly, the method of nearest hierarchical clustering is applied to divide the point cloud if the above cannot be distinguished. It can preferably solve the difficult issue like crossing or overlapping pieces of baggage. Finally, the point cloud is projected to the horizontal plane. By rotating point cloud based on the centre, its minimum bounding rectangle (MBR is obtained. The length and width of luggage are got form MBR. Many experiments in different cases have been done to verify the effectiveness of the algorithm.

  4. Biomolecule-Assisted Hydrothermal Synthesis and Self-Assembly of Bi2Te3 Nanostring-Cluster Hierarchical Structure

    DEFF Research Database (Denmark)

    Mi, Jianli; Lock, Nina; Sun, Ting

    2010-01-01

    A simple biomolecule-assisted hydrothermal approach has been developed for the fabrication of Bi2Te3 thermoelectric nanomaterials. The product has a nanostring-cluster hierarchical structure which is composed of ordered and aligned platelet-like crystals. The platelets are100 nm in diameter...

  5. Classification and unsupervised clustering of LIGO data with Deep Transfer Learning

    Science.gov (United States)

    George, Daniel; Shen, Hongyu; Huerta, E. A.

    2018-05-01

    Gravitational wave detection requires a detailed understanding of the response of the LIGO and Virgo detectors to true signals in the presence of environmental and instrumental noise. Of particular interest is the study of anomalous non-Gaussian transients, such as glitches, since their occurrence rate in LIGO and Virgo data can obscure or even mimic true gravitational wave signals. Therefore, successfully identifying and excising these anomalies from gravitational wave data is of utmost importance for the detection and characterization of true signals and for the accurate computation of their significance. To facilitate this work, we present the first application of deep learning combined with transfer learning to show that knowledge from pretrained models for real-world object recognition can be transferred for classifying spectrograms of glitches. To showcase this new method, we use a data set of twenty-two classes of glitches, curated and labeled by the Gravity Spy project using data collected during LIGO's first discovery campaign. We demonstrate that our Deep Transfer Learning method enables an optimal use of very deep convolutional neural networks for glitch classification given small and unbalanced training data sets, significantly reduces the training time, and achieves state-of-the-art accuracy above 98.8%, lowering the previous error rate by over 60%. More importantly, once trained via transfer learning on the known classes, we show that our neural networks can be truncated and used as feature extractors for unsupervised clustering to automatically group together new unknown classes of glitches and anomalous signals. This novel capability is of paramount importance to identify and remove new types of glitches which will occur as the LIGO/Virgo detectors gradually attain design sensitivity.

  6. Noise-robust unsupervised spike sorting based on discriminative subspace learning with outlier handling.

    Science.gov (United States)

    Keshtkaran, Mohammad Reza; Yang, Zhi

    2017-06-01

    Spike sorting is a fundamental preprocessing step for many neuroscience studies which rely on the analysis of spike trains. Most of the feature extraction and dimensionality reduction techniques that have been used for spike sorting give a projection subspace which is not necessarily the most discriminative one. Therefore, the clusters which appear inherently separable in some discriminative subspace may overlap if projected using conventional feature extraction approaches leading to a poor sorting accuracy especially when the noise level is high. In this paper, we propose a noise-robust and unsupervised spike sorting algorithm based on learning discriminative spike features for clustering. The proposed algorithm uses discriminative subspace learning to extract low dimensional and most discriminative features from the spike waveforms and perform clustering with automatic detection of the number of the clusters. The core part of the algorithm involves iterative subspace selection using linear discriminant analysis and clustering using Gaussian mixture model with outlier detection. A statistical test in the discriminative subspace is proposed to automatically detect the number of the clusters. Comparative results on publicly available simulated and real in vivo datasets demonstrate that our algorithm achieves substantially improved cluster distinction leading to higher sorting accuracy and more reliable detection of clusters which are highly overlapping and not detectable using conventional feature extraction techniques such as principal component analysis or wavelets. By providing more accurate information about the activity of more number of individual neurons with high robustness to neural noise and outliers, the proposed unsupervised spike sorting algorithm facilitates more detailed and accurate analysis of single- and multi-unit activities in neuroscience and brain machine interface studies.

  7. Noise-robust unsupervised spike sorting based on discriminative subspace learning with outlier handling

    Science.gov (United States)

    Keshtkaran, Mohammad Reza; Yang, Zhi

    2017-06-01

    Objective. Spike sorting is a fundamental preprocessing step for many neuroscience studies which rely on the analysis of spike trains. Most of the feature extraction and dimensionality reduction techniques that have been used for spike sorting give a projection subspace which is not necessarily the most discriminative one. Therefore, the clusters which appear inherently separable in some discriminative subspace may overlap if projected using conventional feature extraction approaches leading to a poor sorting accuracy especially when the noise level is high. In this paper, we propose a noise-robust and unsupervised spike sorting algorithm based on learning discriminative spike features for clustering. Approach. The proposed algorithm uses discriminative subspace learning to extract low dimensional and most discriminative features from the spike waveforms and perform clustering with automatic detection of the number of the clusters. The core part of the algorithm involves iterative subspace selection using linear discriminant analysis and clustering using Gaussian mixture model with outlier detection. A statistical test in the discriminative subspace is proposed to automatically detect the number of the clusters. Main results. Comparative results on publicly available simulated and real in vivo datasets demonstrate that our algorithm achieves substantially improved cluster distinction leading to higher sorting accuracy and more reliable detection of clusters which are highly overlapping and not detectable using conventional feature extraction techniques such as principal component analysis or wavelets. Significance. By providing more accurate information about the activity of more number of individual neurons with high robustness to neural noise and outliers, the proposed unsupervised spike sorting algorithm facilitates more detailed and accurate analysis of single- and multi-unit activities in neuroscience and brain machine interface studies.

  8. Enhancement of Tropical Land Cover Mapping with Wavelet-Based Fusion and Unsupervised Clustering of SAR and Landsat Image Data

    Science.gov (United States)

    LeMoigne, Jacqueline; Laporte, Nadine; Netanyahuy, Nathan S.; Zukor, Dorothy (Technical Monitor)

    2001-01-01

    The characterization and the mapping of land cover/land use of forest areas, such as the Central African rainforest, is a very complex task. This complexity is mainly due to the extent of such areas and, as a consequence, to the lack of full and continuous cloud-free coverage of those large regions by one single remote sensing instrument, In order to provide improved vegetation maps of Central Africa and to develop forest monitoring techniques for applications at the local and regional scales, we propose to utilize multi-sensor remote sensing observations coupled with in-situ data. Fusion and clustering of multi-sensor data are the first steps towards the development of such a forest monitoring system. In this paper, we will describe some preliminary experiments involving the fusion of SAR and Landsat image data of the Lope Reserve in Gabon. Similarly to previous fusion studies, our fusion method is wavelet-based. The fusion provides a new image data set which contains more detailed texture features and preserves the large homogeneous regions that are observed by the Thematic Mapper sensor. The fusion step is followed by unsupervised clustering and provides a vegetation map of the area.

  9. A supplier selection using a hybrid grey based hierarchical clustering and artificial bee colony

    Directory of Open Access Journals (Sweden)

    Farshad Faezy Razi

    2014-06-01

    Full Text Available Selection of one or a combination of the most suitable potential providers and outsourcing problem is the most important strategies in logistics and supply chain management. In this paper, selection of an optimal combination of suppliers in inventory and supply chain management are studied and analyzed via multiple attribute decision making approach, data mining and evolutionary optimization algorithms. For supplier selection in supply chain, hierarchical clustering according to the studied indexes first clusters suppliers. Then, according to its cluster, each supplier is evaluated through Grey Relational Analysis. Then the combination of suppliers’ Pareto optimal rank and costs are obtained using Artificial Bee Colony meta-heuristic algorithm. A case study is conducted for a better description of a new algorithm to select a multiple source of suppliers.

  10. Bayesian feature weighting for unsupervised learning, with application to object recognition

    OpenAIRE

    Carbonetto , Peter; De Freitas , Nando; Gustafson , Paul; Thompson , Natalie

    2003-01-01

    International audience; We present a method for variable selection/weighting in an unsupervised learning context using Bayesian shrinkage. The basis for the model parameters and cluster assignments can be computed simultaneous using an efficient EM algorithm. Applying our Bayesian shrinkage model to a complex problem in object recognition (Duygulu, Barnard, de Freitas and Forsyth 2002), our experiments yied good results.

  11. A study of hierarchical clustering of galaxies in an expanding universe

    Science.gov (United States)

    Porter, D. H.

    The nonlinear hierarchical clustering of galaxies in an Einstein-deSitter (Omega = 1), initially white noise mass fluctuations (n = 0) model universe is investigated and shown to be in contradiction with previous results. The model is done in terms of an 11,000-body numerical simulation. The independent statics of 0.72 million particles are used to simulte the boundary conditions. A new method for integrating the Newtonian N-body gravity equations, which has controllable accuracy, incorporates a recursive center of mass reduction, and regularizes two body encounters is used to do the simulation. The coordinate system used here is well suited for the investigation of galaxy clustering, incorporating the independent positions and velocities of an arbitrary number of particles into a logarithmic hierarchy of center of mass nodes. The boundary for the simulation is created by using this hierarchy to map the independent statics of 0.72 million particles into just 4,000 particles. This method for simulating the boundary conditions also has controllable accuracy.

  12. Evaluating Mixture Modeling for Clustering: Recommendations and Cautions

    Science.gov (United States)

    Steinley, Douglas; Brusco, Michael J.

    2011-01-01

    This article provides a large-scale investigation into several of the properties of mixture-model clustering techniques (also referred to as latent class cluster analysis, latent profile analysis, model-based clustering, probabilistic clustering, Bayesian classification, unsupervised learning, and finite mixture models; see Vermunt & Magdison,…

  13. The composite sequential clustering technique for analysis of multispectral scanner data

    Science.gov (United States)

    Su, M. Y.

    1972-01-01

    The clustering technique consists of two parts: (1) a sequential statistical clustering which is essentially a sequential variance analysis, and (2) a generalized K-means clustering. In this composite clustering technique, the output of (1) is a set of initial clusters which are input to (2) for further improvement by an iterative scheme. This unsupervised composite technique was employed for automatic classification of two sets of remote multispectral earth resource observations. The classification accuracy by the unsupervised technique is found to be comparable to that by traditional supervised maximum likelihood classification techniques. The mathematical algorithms for the composite sequential clustering program and a detailed computer program description with job setup are given.

  14. Chemical modeling of groundwater in the Banat Plain, southwestern Romania, with elevated As content and co-occurring species by combining diagrams and unsupervised multivariate statistical approaches.

    Science.gov (United States)

    Butaciu, Sinziana; Senila, Marin; Sarbu, Costel; Ponta, Michaela; Tanaselia, Claudiu; Cadar, Oana; Roman, Marius; Radu, Emil; Sima, Mihaela; Frentiu, Tiberiu

    2017-04-01

    The study proposes a combined model based on diagrams (Gibbs, Piper, Stuyfzand Hydrogeochemical Classification System) and unsupervised statistical approaches (Cluster Analysis, Principal Component Analysis, Fuzzy Principal Component Analysis, Fuzzy Hierarchical Cross-Clustering) to describe natural enrichment of inorganic arsenic and co-occurring species in groundwater in the Banat Plain, southwestern Romania. Speciation of inorganic As (arsenite, arsenate), ion concentrations (Na + , K + , Ca 2+ , Mg 2+ , HCO 3 - , Cl - , F - , SO 4 2- , PO 4 3- , NO 3 - ), pH, redox potential, conductivity and total dissolved substances were performed. Classical diagrams provided the hydrochemical characterization, while statistical approaches were helpful to establish (i) the mechanism of naturally occurring of As and F - species and the anthropogenic one for NO 3 - , SO 4 2- , PO 4 3- and K + and (ii) classification of groundwater based on content of arsenic species. The HCO 3 - type of local groundwater and alkaline pH (8.31-8.49) were found to be responsible for the enrichment of arsenic species and occurrence of F - but by different paths. The PO 4 3- -AsO 4 3- ion exchange, water-rock interaction (silicates hydrolysis and desorption from clay) were associated to arsenate enrichment in the oxidizing aquifer. Fuzzy Hierarchical Cross-Clustering was the strongest tool for the rapid simultaneous classification of groundwaters as a function of arsenic content and hydrogeochemical characteristics. The approach indicated the Na + -F - -pH cluster as marker for groundwater with naturally elevated As and highlighted which parameters need to be monitored. A chemical conceptual model illustrating the natural and anthropogenic paths and enrichment of As and co-occurring species in the local groundwater supported by mineralogical analysis of rocks was established. Copyright © 2016 Elsevier Ltd. All rights reserved.

  15. A hierarchical clustering scheme approach to assessment of IP-network traffic using detrended fluctuation analysis

    Science.gov (United States)

    Takuma, Takehisa; Masugi, Masao

    2009-03-01

    This paper presents an approach to the assessment of IP-network traffic in terms of the time variation of self-similarity. To get a comprehensive view in analyzing the degree of long-range dependence (LRD) of IP-network traffic, we use a hierarchical clustering scheme, which provides a way to classify high-dimensional data with a tree-like structure. Also, in the LRD-based analysis, we employ detrended fluctuation analysis (DFA), which is applicable to the analysis of long-range power-law correlations or LRD in non-stationary time-series signals. Based on sequential measurements of IP-network traffic at two locations, this paper derives corresponding values for the LRD-related parameter α that reflects the degree of LRD of measured data. In performing the hierarchical clustering scheme, we use three parameters: the α value, average throughput, and the proportion of network traffic that exceeds 80% of network bandwidth for each measured data set. We visually confirm that the traffic data can be classified in accordance with the network traffic properties, resulting in that the combined depiction of the LRD and other factors can give us an effective assessment of network conditions at different times.

  16. Implementation of hierarchical clustering using k-mer sparse matrix to analyze MERS-CoV genetic relationship

    Science.gov (United States)

    Bustamam, A.; Ulul, E. D.; Hura, H. F. A.; Siswantining, T.

    2017-07-01

    Hierarchical clustering is one of effective methods in creating a phylogenetic tree based on the distance matrix between DNA (deoxyribonucleic acid) sequences. One of the well-known methods to calculate the distance matrix is k-mer method. Generally, k-mer is more efficient than some distance matrix calculation techniques. The steps of k-mer method are started from creating k-mer sparse matrix, and followed by creating k-mer singular value vectors. The last step is computing the distance amongst vectors. In this paper, we analyze the sequences of MERS-CoV (Middle East Respiratory Syndrome - Coronavirus) DNA by implementing hierarchical clustering using k-mer sparse matrix in order to perform the phylogenetic analysis. Our results show that the ancestor of our MERS-CoV is coming from Egypt. Moreover, we found that the MERS-CoV infection that occurs in one country may not necessarily come from the same country of origin. This suggests that the process of MERS-CoV mutation might not only be influenced by geographical factor.

  17. Unsupervised consensus cluster analysis of [18F]-fluoroethyl-L-tyrosine positron emission tomography identified textural features for the diagnosis of pseudoprogression in high-grade glioma.

    Science.gov (United States)

    Kebir, Sied; Khurshid, Zain; Gaertner, Florian C; Essler, Markus; Hattingen, Elke; Fimmers, Rolf; Scheffler, Björn; Herrlinger, Ulrich; Bundschuh, Ralph A; Glas, Martin

    2017-01-31

    Timely detection of pseudoprogression (PSP) is crucial for the management of patients with high-grade glioma (HGG) but remains difficult. Textural features of O-(2-[18F]fluoroethyl)-L-tyrosine positron emission tomography (FET-PET) mirror tumor uptake heterogeneity; some of them may be associated with tumor progression. Fourteen patients with HGG and suspected of PSP underwent FET-PET imaging. A set of 19 conventional and textural FET-PET features were evaluated and subjected to unsupervised consensus clustering. The final diagnosis of true progression vs. PSP was based on follow-up MRI using RANO criteria. Three robust clusters have been identified based on 10 predominantly textural FET-PET features. None of the patients with PSP fell into cluster 2, which was associated with high values for textural FET-PET markers of uptake heterogeneity. Three out of 4 patients with PSP were assigned to cluster 3 that was largely associated with low values of textural FET-PET features. By comparison, tumor-to-normal brain ratio (TNRmax) at the optimal cutoff 2.1 was less predictive of PSP (negative predictive value 57% for detecting true progression, p=0.07 vs. 75% with cluster 3, p=0.04). Clustering based on textural O-(2-[18F]fluoroethyl)-L-tyrosine PET features may provide valuable information in assessing the elusive phenomenon of pseudoprogression.

  18. Symptom Clusters in People Living with HIV Attending Five Palliative Care Facilities in Two Sub-Saharan African Countries: A Hierarchical Cluster Analysis.

    Science.gov (United States)

    Moens, Katrien; Siegert, Richard J; Taylor, Steve; Namisango, Eve; Harding, Richard

    2015-01-01

    Symptom research across conditions has historically focused on single symptoms, and the burden of multiple symptoms and their interactions has been relatively neglected especially in people living with HIV. Symptom cluster studies are required to set priorities in treatment planning, and to lessen the total symptom burden. This study aimed to identify and compare symptom clusters among people living with HIV attending five palliative care facilities in two sub-Saharan African countries. Data from cross-sectional self-report of seven-day symptom prevalence on the 32-item Memorial Symptom Assessment Scale-Short Form were used. A hierarchical cluster analysis was conducted using Ward's method applying squared Euclidean Distance as the similarity measure to determine the clusters. Contingency tables, X2 tests and ANOVA were used to compare the clusters by patient specific characteristics and distress scores. Among the sample (N=217) the mean age was 36.5 (SD 9.0), 73.2% were female, and 49.1% were on antiretroviral therapy (ART). The cluster analysis produced five symptom clusters identified as: 1) dermatological; 2) generalised anxiety and elimination; 3) social and image; 4) persistently present; and 5) a gastrointestinal-related symptom cluster. The patients in the first three symptom clusters reported the highest physical and psychological distress scores. Patient characteristics varied significantly across the five clusters by functional status (worst functional physical status in cluster one, ppeople living with HIV with longitudinally collected symptom data to test cluster stability and identify common symptom trajectories is recommended.

  19. Statistical mechanics of semi-supervised clustering in sparse graphs

    International Nuclear Information System (INIS)

    Ver Steeg, Greg; Galstyan, Aram; Allahverdyan, Armen E

    2011-01-01

    We theoretically study semi-supervised clustering in sparse graphs in the presence of pair-wise constraints on the cluster assignments of nodes. We focus on bi-cluster graphs and study the impact of semi-supervision for varying constraint density and overlap between the clusters. Recent results for unsupervised clustering in sparse graphs indicate that there is a critical ratio of within-cluster and between-cluster connectivities below which clusters cannot be recovered with better than random accuracy. The goal of this paper is to examine the impact of pair-wise constraints on the clustering accuracy. Our results suggest that the addition of constraints does not provide automatic improvement over the unsupervised case. When the density of the constraints is sufficiently small, their only impact is to shift the detection threshold while preserving the criticality. Conversely, if the density of (hard) constraints is above the percolation threshold, the criticality is suppressed and the detection threshold disappears

  20. K-means clustering for optimal partitioning and dynamic load balancing of parallel hierarchical N-body simulations

    International Nuclear Information System (INIS)

    Marzouk, Youssef M.; Ghoniem, Ahmed F.

    2005-01-01

    A number of complex physical problems can be approached through N-body simulation, from fluid flow at high Reynolds number to gravitational astrophysics and molecular dynamics. In all these applications, direct summation is prohibitively expensive for large N and thus hierarchical methods are employed for fast summation. This work introduces new algorithms, based on k-means clustering, for partitioning parallel hierarchical N-body interactions. We demonstrate that the number of particle-cluster interactions and the order at which they are performed are directly affected by partition geometry. Weighted k-means partitions minimize the sum of clusters' second moments and create well-localized domains, and thus reduce the computational cost of N-body approximations by enabling the use of lower-order approximations and fewer cells. We also introduce compatible techniques for dynamic load balancing, including adaptive scaling of cluster volumes and adaptive redistribution of cluster centroids. We demonstrate the performance of these algorithms by constructing a parallel treecode for vortex particle simulations, based on the serial variable-order Cartesian code developed by Lindsay and Krasny [Journal of Computational Physics 172 (2) (2001) 879-907]. The method is applied to vortex simulations of a transverse jet. Results show outstanding parallel efficiencies even at high concurrencies, with velocity evaluation errors maintained at or below their serial values; on a realistic distribution of 1.2 million vortex particles, we observe a parallel efficiency of 98% on 1024 processors. Excellent load balance is achieved even in the face of several obstacles, such as an irregular, time-evolving particle distribution containing a range of length scales and the continual introduction of new vortex particles throughout the domain. Moreover, results suggest that k-means yields a more efficient partition of the domain than a global oct-tree

  1. Single pass kernel k-means clustering method

    Indian Academy of Sciences (India)

    In unsupervised classification, kernel -means clustering method has been shown to perform better than conventional -means clustering method in ... 518501, India; Department of Computer Science and Engineering, Jawaharlal Nehru Technological University, Anantapur College of Engineering, Anantapur 515002, India ...

  2. A comparison of hierarchical cluster analysis and league table rankings as methods for analysis and presentation of district health system performance data in Uganda.

    Science.gov (United States)

    Tashobya, Christine K; Dubourg, Dominique; Ssengooba, Freddie; Speybroeck, Niko; Macq, Jean; Criel, Bart

    2016-03-01

    In 2003, the Uganda Ministry of Health introduced the district league table for district health system performance assessment. The league table presents district performance against a number of input, process and output indicators and a composite index to rank districts. This study explores the use of hierarchical cluster analysis for analysing and presenting district health systems performance data and compares this approach with the use of the league table in Uganda. Ministry of Health and district plans and reports, and published documents were used to provide information on the development and utilization of the Uganda district league table. Quantitative data were accessed from the Ministry of Health databases. Statistical analysis using SPSS version 20 and hierarchical cluster analysis, utilizing Wards' method was used. The hierarchical cluster analysis was conducted on the basis of seven clusters determined for each year from 2003 to 2010, ranging from a cluster of good through moderate-to-poor performers. The characteristics and membership of clusters varied from year to year and were determined by the identity and magnitude of performance of the individual variables. Criticisms of the league table include: perceived unfairness, as it did not take into consideration district peculiarities; and being oversummarized and not adequately informative. Clustering organizes the many data points into clusters of similar entities according to an agreed set of indicators and can provide the beginning point for identifying factors behind the observed performance of districts. Although league table ranking emphasize summation and external control, clustering has the potential to encourage a formative, learning approach. More research is required to shed more light on factors behind observed performance of the different clusters. Other countries especially low-income countries that share many similarities with Uganda can learn from these experiences. © The Author 2015

  3. Prioritizing the risk of plant pests by clustering methods; self-organising maps, k-means and hierarchical clustering

    Directory of Open Access Journals (Sweden)

    Susan Worner

    2013-09-01

    -means, hierarchical clustering and the incorporation of the SOM analysis into criteria based approaches to assess pest risk.

  4. Characterizing the course of back pain after osteoporotic vertebral fracture: a hierarchical cluster analysis of a prospective cohort study.

    Science.gov (United States)

    Toyoda, Hiromitsu; Takahashi, Shinji; Hoshino, Masatoshi; Takayama, Kazushi; Iseki, Kazumichi; Sasaoka, Ryuichi; Tsujio, Tadao; Yasuda, Hiroyuki; Sasaki, Takeharu; Kanematsu, Fumiaki; Kono, Hiroshi; Nakamura, Hiroaki

    2017-09-23

    This study demonstrated four distinct patterns in the course of back pain after osteoporotic vertebral fracture (OVF). Greater angular instability in the first 6 months after the baseline was one factor affecting back pain after OVF. Understanding the natural course of symptomatic acute OVF is important in deciding the optimal treatment strategy. We used latent class analysis to classify the course of back pain after OVF and identify the risk factors associated with persistent pain. This multicenter cohort study included 218 consecutive patients with ≤ 2-week-old OVFs who were enrolled at 11 institutions. Dynamic x-rays and back pain assessment with a visual analog scale (VAS) were obtained at enrollment and at 1-, 3-, and 6-month follow-ups. The VAS scores were used to characterize patient groups, using hierarchical cluster analysis. VAS for 128 patients was used for hierarchical cluster analysis. Analysis yielded four clusters representing different patterns of back pain progression. Cluster 1 patients (50.8%) had stable, mild pain. Cluster 2 patients (21.1%) started with moderate pain and progressed quickly to very low pain. Patients in cluster 3 (10.9%) had moderate pain that initially improved but worsened after 3 months. Cluster 4 patients (17.2%) had persistent severe pain. Patients in cluster 4 showed significant high baseline pain intensity, higher degree of angular instability, and higher number of previous OVFs, and tended to lack regular exercise. In contrast, patients in cluster 2 had significantly lower baseline VAS and less angular instability. We identified four distinct groups of OVF patients with different patterns of back pain progression. Understanding the course of back pain after OVF may help in its management and contribute to future treatment trials.

  5. A Resting-State Brain Functional Network Study in MDD Based on Minimum Spanning Tree Analysis and the Hierarchical Clustering

    Directory of Open Access Journals (Sweden)

    Xiaowei Li

    2017-01-01

    Full Text Available A large number of studies demonstrated that major depressive disorder (MDD is characterized by the alterations in brain functional connections which is also identifiable during the brain’s “resting-state.” But, in the present study, the approach of constructing functional connectivity is often biased by the choice of the threshold. Besides, more attention was paid to the number and length of links in brain networks, and the clustering partitioning of nodes was unclear. Therefore, minimum spanning tree (MST analysis and the hierarchical clustering were first used for the depression disease in this study. Resting-state electroencephalogram (EEG sources were assessed from 15 healthy and 23 major depressive subjects. Then the coherence, MST, and the hierarchical clustering were obtained. In the theta band, coherence analysis showed that the EEG coherence of the MDD patients was significantly higher than that of the healthy controls especially in the left temporal region. The MST results indicated the higher leaf fraction in the depressed group. Compared with the normal group, the major depressive patients lost clustering in frontal regions. Our findings suggested that there was a stronger brain interaction in the MDD group and a left-right functional imbalance in the frontal regions for MDD controls.

  6. Assessment of anaesthetic depth by clustering analysis and autoregressive modelling of electroencephalograms

    DEFF Research Database (Denmark)

    Thomsen, C E; Rosenfalck, A; Nørregaard Christensen, K

    1991-01-01

    The brain activity electroencephalogram (EEG) was recorded from 30 healthy women scheduled for hysterectomy. The patients were anaesthetized with isoflurane, halothane or etomidate/fentanyl. A multiparametric method was used for extraction of amplitude and frequency information from the EEG....... The method applied autoregressive modelling of the signal, segmented in 2 s fixed intervals. The features from the EEG segments were used for learning and for classification. The learning process was unsupervised and hierarchical clustering analysis was used to construct a learning set of EEG amplitude......-frequency patterns for each of the three anaesthetic drugs. These EEG patterns were assigned to a colour code corresponding to similar clinical states. A common learning set could be used for all patients anaesthetized with the same drug. The classification process could be performed on-line and the results were...

  7. Predicting protein complexes from weighted protein-protein interaction graphs with a novel unsupervised methodology: Evolutionary enhanced Markov clustering.

    Science.gov (United States)

    Theofilatos, Konstantinos; Pavlopoulou, Niki; Papasavvas, Christoforos; Likothanassis, Spiros; Dimitrakopoulos, Christos; Georgopoulos, Efstratios; Moschopoulos, Charalampos; Mavroudi, Seferina

    2015-03-01

    Proteins are considered to be the most important individual components of biological systems and they combine to form physical protein complexes which are responsible for certain molecular functions. Despite the large availability of protein-protein interaction (PPI) information, not much information is available about protein complexes. Experimental methods are limited in terms of time, efficiency, cost and performance constraints. Existing computational methods have provided encouraging preliminary results, but they phase certain disadvantages as they require parameter tuning, some of them cannot handle weighted PPI data and others do not allow a protein to participate in more than one protein complex. In the present paper, we propose a new fully unsupervised methodology for predicting protein complexes from weighted PPI graphs. The proposed methodology is called evolutionary enhanced Markov clustering (EE-MC) and it is a hybrid combination of an adaptive evolutionary algorithm and a state-of-the-art clustering algorithm named enhanced Markov clustering. EE-MC was compared with state-of-the-art methodologies when applied to datasets from the human and the yeast Saccharomyces cerevisiae organisms. Using public available datasets, EE-MC outperformed existing methodologies (in some datasets the separation metric was increased by 10-20%). Moreover, when applied to new human datasets its performance was encouraging in the prediction of protein complexes which consist of proteins with high functional similarity. In specific, 5737 protein complexes were predicted and 72.58% of them are enriched for at least one gene ontology (GO) function term. EE-MC is by design able to overcome intrinsic limitations of existing methodologies such as their inability to handle weighted PPI networks, their constraint to assign every protein in exactly one cluster and the difficulties they face concerning the parameter tuning. This fact was experimentally validated and moreover, new

  8. Using hierarchical clustering methods to classify motor activities of COPD patients from wearable sensor data

    Directory of Open Access Journals (Sweden)

    Reilly John J

    2005-06-01

    Full Text Available Abstract Background Advances in miniature sensor technology have led to the development of wearable systems that allow one to monitor motor activities in the field. A variety of classifiers have been proposed in the past, but little has been done toward developing systematic approaches to assess the feasibility of discriminating the motor tasks of interest and to guide the choice of the classifier architecture. Methods A technique is introduced to address this problem according to a hierarchical framework and its use is demonstrated for the application of detecting motor activities in patients with chronic obstructive pulmonary disease (COPD undergoing pulmonary rehabilitation. Accelerometers were used to collect data for 10 different classes of activity. Features were extracted to capture essential properties of the data set and reduce the dimensionality of the problem at hand. Cluster measures were utilized to find natural groupings in the data set and then construct a hierarchy of the relationships between clusters to guide the process of merging clusters that are too similar to distinguish reliably. It provides a means to assess whether the benefits of merging for performance of a classifier outweigh the loss of resolution incurred through merging. Results Analysis of the COPD data set demonstrated that motor tasks related to ambulation can be reliably discriminated from tasks performed in a seated position with the legs in motion or stationary using two features derived from one accelerometer. Classifying motor tasks within the category of activities related to ambulation requires more advanced techniques. While in certain cases all the tasks could be accurately classified, in others merging clusters associated with different motor tasks was necessary. When merging clusters, it was found that the proposed method could lead to more than 12% improvement in classifier accuracy while retaining resolution of 4 tasks. Conclusion Hierarchical

  9. On hierarchical solutions to the BBGKY hierarchy

    Science.gov (United States)

    Hamilton, A. J. S.

    1988-01-01

    It is thought that the gravitational clustering of galaxies in the universe may approach a scale-invariant, hierarchical form in the small separation, large-clustering regime. Past attempts to solve the Born-Bogoliubov-Green-Kirkwood-Yvon (BBGKY) hierarchy in this regime have assumed a certain separable hierarchical form for the higher order correlation functions of galaxies in phase space. It is shown here that such separable solutions to the BBGKY equations must satisfy the condition that the clustered component of the solution has cluster-cluster correlations equal to galaxy-galaxy correlations to all orders. The solutions also admit the presence of an arbitrary unclustered component, which plays no dyamical role in the large-clustering regime. These results are a particular property of the specific separable model assumed for the correlation functions in phase space, not an intrinsic property of spatially hierarchical solutions to the BBGKY hierarchy. The observed distribution of galaxies does not satisfy the required conditions. The disagreement between theory and observation may be traced, at least in part, to initial conditions which, if Gaussian, already have cluster correlations greater than galaxy correlations.

  10. Searching remote homology with spectral clustering with symmetry in neighborhood cluster kernels.

    Directory of Open Access Journals (Sweden)

    Ujjwal Maulik

    Full Text Available Remote homology detection among proteins utilizing only the unlabelled sequences is a central problem in comparative genomics. The existing cluster kernel methods based on neighborhoods and profiles and the Markov clustering algorithms are currently the most popular methods for protein family recognition. The deviation from random walks with inflation or dependency on hard threshold in similarity measure in those methods requires an enhancement for homology detection among multi-domain proteins. We propose to combine spectral clustering with neighborhood kernels in Markov similarity for enhancing sensitivity in detecting homology independent of "recent" paralogs. The spectral clustering approach with new combined local alignment kernels more effectively exploits the unsupervised protein sequences globally reducing inter-cluster walks. When combined with the corrections based on modified symmetry based proximity norm deemphasizing outliers, the technique proposed in this article outperforms other state-of-the-art cluster kernels among all twelve implemented kernels. The comparison with the state-of-the-art string and mismatch kernels also show the superior performance scores provided by the proposed kernels. Similar performance improvement also is found over an existing large dataset. Therefore the proposed spectral clustering framework over combined local alignment kernels with modified symmetry based correction achieves superior performance for unsupervised remote homolog detection even in multi-domain and promiscuous domain proteins from Genolevures database families with better biological relevance. Source code available upon request.sarkar@labri.fr.

  11. Hierarchical Cluster Analysis of Semicircular Canal and Otolith Deficits in Bilateral Vestibulopathy

    Directory of Open Access Journals (Sweden)

    Alexander A. Tarnutzer

    2018-04-01

    Full Text Available BackgroundGait imbalance and oscillopsia are frequent complaints of bilateral vestibular loss (BLV. Video-head-impulse testing (vHIT of all six semicircular canals (SCCs has demonstrated varying involvement of the different canals. Sparing of anterior-canal function has been linked to aminoglycoside-related vestibulopathy and Menière’s disease. We hypothesized that utricular and saccular impairment [assessed by vestibular-evoked myogenic potentials (VEMPs] may be disease-specific also, possibly facilitating the differential diagnosis.MethodsWe searched our vHIT database (n = 3,271 for patients with bilaterally impaired SCC function who also received ocular VEMPs (oVEMPs and cervical VEMPs (cVEMPs and identified 101 patients. oVEMP/cVEMP latencies above the 95th percentile and peak-to-peak amplitudes below the 5th percentile of normal were considered abnormal. Frequency of impairment of vestibular end organs (horizontal/anterior/posterior SCC, utriculus/sacculus was analyzed with hierarchical cluster analysis and correlated with the underlying etiology.ResultsRates of utricular and saccular loss of function were similar (87.1 vs. 78.2%, p = 0.136, Fisher’s exact test. oVEMP abnormalities were found more frequent in aminoglycoside-related bilateral vestibular loss (BVL compared with Menière’s disease (91.7 vs. 54.6%, p = 0.039. Hierarchical cluster analysis indicated distinct patterns of vestibular end-organ impairment, showing that the results for the same end-organs on both sides are more similar than to other end-organs. Relative sparing of anterior-canal function was reflected in late merging with the other end-organs, emphasizing their distinct state. An anatomically corresponding pattern of SCC/otolith hypofunction was present in 60.4% (oVEMPs vs. horizontal SCCs, 34.7% (oVEMPs vs. anterior SCCs, and 48.5% (cVEMPs vs. posterior SCCs of cases. Average (±1 SD number of damaged sensors was 6.8 ± 2.2 out of 10

  12. Efficient visible light photocatalytic NO{sub x} removal with cationic Ag clusters-grafted (BiO){sub 2}CO{sub 3} hierarchical superstructures

    Energy Technology Data Exchange (ETDEWEB)

    Feng, Xin [Chongqing Key Laboratory of Catalysis and Functional Organic Molecules, College of Environment and Resources, Engineering Research Center for Waste Oil Recovery Technology and Equipment of Ministry of Education, College of Environment and Resources, Chongqing Technology and Business University, Chongqing 40067 (China); Zhang, Wendong [Department of Scientific Research Management, Chongqing Normal University, Chongqing 401331 (China); Deng, Hua [State Key Joint Laboratory of Environment Simulation and Pollution Control, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085 (China); Ni, Zilin [Department of Scientific Research Management, Chongqing Normal University, Chongqing 401331 (China); Dong, Fan, E-mail: dfctbu@126.com [Chongqing Key Laboratory of Catalysis and Functional Organic Molecules, College of Environment and Resources, Engineering Research Center for Waste Oil Recovery Technology and Equipment of Ministry of Education, College of Environment and Resources, Chongqing Technology and Business University, Chongqing 40067 (China); Zhang, Yuxin, E-mail: zhangyuxin@cqu.edu.cn [College of Materials Science and Engineering, National Key Laboratory of Fundamental Science of Micro/Nano-Devices and System Technology, Chongqing University, Chongqing 400044 (China)

    2017-01-15

    Graphical abstract: The cationic Ag clusters-grafted (BiO){sub 2}CO{sub 3} hierarchical superstructures exhibits highly enhanced visible light photocatalytic air purification through an interfacial charge transfer process induced by Ag clusters. - Highlights: • Microstructural optimization and surface cluster-grafting were firstly combined. • Cationic Ag clusters were grafted on the surface of (BiO){sub 2}CO{sub 3} superstructures. • The Ag clusters-grafted BHS displayed enhanced visible light photocatalysis. • Direct interfacial charge transfer (IFCT) from BHS to Ag clusters was proposed. • The charge transfer process and the dominant reactive species were revealed. - Abstract: A facile method was developed to graft cationic Ag clusters on (BiO){sub 2}CO{sub 3} hierarchical superstructures (BHS) surface to improve their visible light activity. Significantly, the resultant Ag clusters-grafted BHS displayed a highly enhanced visible light photocatalytic performance for NOx removal due to the direct interfacial charge transfer (IFCT) from BHS to Ag clusters. The chemical and coordination state of the cationic Ag clusters was determined with the extended X-ray absorption fine structure (EXAFS) and a theoretical structure model was proposed for this unique Ag clusters. The charge transfer process and the dominant reactive species (·OH) were revealed on the basis of electron spin resonance (ESR) trapping. A new photocatalysis mechanism of Ag clusters-grafted BHS under visible light involving IFCT process was uncovered. In addition, the cationic Ag clusters-grafted BHS also demonstrated high photochemical and structural stability under repeated photocatalysis runs. The perspective of enhancing photocatalysis through combination of microstructural optimization and IFCT could provide a new avenue for the developing efficient visible light photocatalysts.

  13. Unsupervised daily routine and activity discovery in smart homes.

    Science.gov (United States)

    Jie Yin; Qing Zhang; Karunanithi, Mohan

    2015-08-01

    The ability to accurately recognize daily activities of residents is a core premise of smart homes to assist with remote health monitoring. Most of the existing methods rely on a supervised model trained from a preselected and manually labeled set of activities, which are often time-consuming and costly to obtain in practice. In contrast, this paper presents an unsupervised method for discovering daily routines and activities for smart home residents. Our proposed method first uses a Markov chain to model a resident's locomotion patterns at different times of day and discover clusters of daily routines at the macro level. For each routine cluster, it then drills down to further discover room-level activities at the micro level. The automatic identification of daily routines and activities is useful for understanding indicators of functional decline of elderly people and suggesting timely interventions.

  14. An automatic taxonomy of galaxy morphology using unsupervised machine learning

    Science.gov (United States)

    Hocking, Alex; Geach, James E.; Sun, Yi; Davey, Neil

    2018-01-01

    We present an unsupervised machine learning technique that automatically segments and labels galaxies in astronomical imaging surveys using only pixel data. Distinct from previous unsupervised machine learning approaches used in astronomy we use no pre-selection or pre-filtering of target galaxy type to identify galaxies that are similar. We demonstrate the technique on the Hubble Space Telescope (HST) Frontier Fields. By training the algorithm using galaxies from one field (Abell 2744) and applying the result to another (MACS 0416.1-2403), we show how the algorithm can cleanly separate early and late type galaxies without any form of pre-directed training for what an 'early' or 'late' type galaxy is. We then apply the technique to the HST Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey (CANDELS) fields, creating a catalogue of approximately 60 000 classifications. We show how the automatic classification groups galaxies of similar morphological (and photometric) type and make the classifications public via a catalogue, a visual catalogue and galaxy similarity search. We compare the CANDELS machine-based classifications to human-classifications from the Galaxy Zoo: CANDELS project. Although there is not a direct mapping between Galaxy Zoo and our hierarchical labelling, we demonstrate a good level of concordance between human and machine classifications. Finally, we show how the technique can be used to identify rarer objects and present lensed galaxy candidates from the CANDELS imaging.

  15. Partitional clustering algorithms

    CERN Document Server

    2015-01-01

    This book summarizes the state-of-the-art in partitional clustering. Clustering, the unsupervised classification of patterns into groups, is one of the most important tasks in exploratory data analysis. Primary goals of clustering include gaining insight into, classifying, and compressing data. Clustering has a long and rich history that spans a variety of scientific disciplines including anthropology, biology, medicine, psychology, statistics, mathematics, engineering, and computer science. As a result, numerous clustering algorithms have been proposed since the early 1950s. Among these algorithms, partitional (nonhierarchical) ones have found many applications, especially in engineering and computer science. This book provides coverage of consensus clustering, constrained clustering, large scale and/or high dimensional clustering, cluster validity, cluster visualization, and applications of clustering. Examines clustering as it applies to large and/or high-dimensional data sets commonly encountered in reali...

  16. Hierarchical ordering with partial pairwise hierarchical relationships on the macaque brain data sets.

    Directory of Open Access Journals (Sweden)

    Woosang Lim

    Full Text Available Hierarchical organizations of information processing in the brain networks have been known to exist and widely studied. To find proper hierarchical structures in the macaque brain, the traditional methods need the entire pairwise hierarchical relationships between cortical areas. In this paper, we present a new method that discovers hierarchical structures of macaque brain networks by using partial information of pairwise hierarchical relationships. Our method uses a graph-based manifold learning to exploit inherent relationship, and computes pseudo distances of hierarchical levels for every pair of cortical areas. Then, we compute hierarchy levels of all cortical areas by minimizing the sum of squared hierarchical distance errors with the hierarchical information of few cortical areas. We evaluate our method on the macaque brain data sets whose true hierarchical levels are known as the FV91 model. The experimental results show that hierarchy levels computed by our method are similar to the FV91 model, and its errors are much smaller than the errors of hierarchical clustering approaches.

  17. K­MEANS CLUSTERING FOR HIDDEN MARKOV MODEL

    NARCIS (Netherlands)

    Perrone, M.P.; Connell, S.D.

    2004-01-01

    An unsupervised k­means clustering algorithm for hidden Markov models is described and applied to the task of generating subclass models for individual handwritten character classes. The algorithm is compared to a related clustering method and shown to give a relative change in the error rate of as

  18. Semisupervised Clustering by Iterative Partition and Regression with Neuroscience Applications

    Directory of Open Access Journals (Sweden)

    Guoqi Qian

    2016-01-01

    Full Text Available Regression clustering is a mixture of unsupervised and supervised statistical learning and data mining method which is found in a wide range of applications including artificial intelligence and neuroscience. It performs unsupervised learning when it clusters the data according to their respective unobserved regression hyperplanes. The method also performs supervised learning when it fits regression hyperplanes to the corresponding data clusters. Applying regression clustering in practice requires means of determining the underlying number of clusters in the data, finding the cluster label of each data point, and estimating the regression coefficients of the model. In this paper, we review the estimation and selection issues in regression clustering with regard to the least squares and robust statistical methods. We also provide a model selection based technique to determine the number of regression clusters underlying the data. We further develop a computing procedure for regression clustering estimation and selection. Finally, simulation studies are presented for assessing the procedure, together with analyzing a real data set on RGB cell marking in neuroscience to illustrate and interpret the method.

  19. Hierarchical video summarization

    Science.gov (United States)

    Ratakonda, Krishna; Sezan, M. Ibrahim; Crinon, Regis J.

    1998-12-01

    We address the problem of key-frame summarization of vide in the absence of any a priori information about its content. This is a common problem that is encountered in home videos. We propose a hierarchical key-frame summarization algorithm where a coarse-to-fine key-frame summary is generated. A hierarchical key-frame summary facilitates multi-level browsing where the user can quickly discover the content of the video by accessing its coarsest but most compact summary and then view a desired segment of the video with increasingly more detail. At the finest level, the summary is generated on the basis of color features of video frames, using an extension of a recently proposed key-frame extraction algorithm. The finest level key-frames are recursively clustered using a novel pairwise K-means clustering approach with temporal consecutiveness constraint. We also address summarization of MPEG-2 compressed video without fully decoding the bitstream. We also propose efficient mechanisms that facilitate decoding the video when the hierarchical summary is utilized in browsing and playback of video segments starting at selected key-frames.

  20. A Clustering Routing Protocol for Mobile Ad Hoc Networks

    Directory of Open Access Journals (Sweden)

    Jinke Huang

    2016-01-01

    Full Text Available The dynamic topology of a mobile ad hoc network poses a real challenge in the design of hierarchical routing protocol, which combines proactive with reactive routing protocols and takes advantages of both. And as an essential technique of hierarchical routing protocol, clustering of nodes provides an efficient method of establishing a hierarchical structure in mobile ad hoc networks. In this paper, we designed a novel clustering algorithm and a corresponding hierarchical routing protocol for large-scale mobile ad hoc networks. Each cluster is composed of a cluster head, several cluster gateway nodes, several cluster guest nodes, and other cluster members. The proposed routing protocol uses proactive protocol between nodes within individual clusters and reactive protocol between clusters. Simulation results show that the proposed clustering algorithm and hierarchical routing protocol provide superior performance with several advantages over existing clustering algorithm and routing protocol, respectively.

  1. Permutation Tests of Hierarchical Cluster Analyses of Carrion Communities and Their Potential Use in Forensic Entomology.

    Science.gov (United States)

    van der Ham, Joris L

    2016-05-19

    Forensic entomologists can use carrion communities' ecological succession data to estimate the postmortem interval (PMI). Permutation tests of hierarchical cluster analyses of these data provide a conceptual method to estimate part of the PMI, the post-colonization interval (post-CI). This multivariate approach produces a baseline of statistically distinct clusters that reflect changes in the carrion community composition during the decomposition process. Carrion community samples of unknown post-CIs are compared with these baseline clusters to estimate the post-CI. In this short communication, I use data from previously published studies to demonstrate the conceptual feasibility of this multivariate approach. Analyses of these data produce series of significantly distinct clusters, which represent carrion communities during 1- to 20-day periods of the decomposition process. For 33 carrion community samples, collected over an 11-day period, this approach correctly estimated the post-CI within an average range of 3.1 days. © The Authors 2016. Published by Oxford University Press on behalf of Entomological Society of America. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

  2. Multichannel biomedical time series clustering via hierarchical probabilistic latent semantic analysis.

    Science.gov (United States)

    Wang, Jin; Sun, Xiangping; Nahavandi, Saeid; Kouzani, Abbas; Wu, Yuchuan; She, Mary

    2014-11-01

    Biomedical time series clustering that automatically groups a collection of time series according to their internal similarity is of importance for medical record management and inspection such as bio-signals archiving and retrieval. In this paper, a novel framework that automatically groups a set of unlabelled multichannel biomedical time series according to their internal structural similarity is proposed. Specifically, we treat a multichannel biomedical time series as a document and extract local segments from the time series as words. We extend a topic model, i.e., the Hierarchical probabilistic Latent Semantic Analysis (H-pLSA), which was originally developed for visual motion analysis to cluster a set of unlabelled multichannel time series. The H-pLSA models each channel of the multichannel time series using a local pLSA in the first layer. The topics learned in the local pLSA are then fed to a global pLSA in the second layer to discover the categories of multichannel time series. Experiments on a dataset extracted from multichannel Electrocardiography (ECG) signals demonstrate that the proposed method performs better than previous state-of-the-art approaches and is relatively robust to the variations of parameters including length of local segments and dictionary size. Although the experimental evaluation used the multichannel ECG signals in a biometric scenario, the proposed algorithm is a universal framework for multichannel biomedical time series clustering according to their structural similarity, which has many applications in biomedical time series management. Copyright © 2014 Elsevier Ireland Ltd. All rights reserved.

  3. Efficient algorithms for accurate hierarchical clustering of huge datasets: tackling the entire protein space.

    Science.gov (United States)

    Loewenstein, Yaniv; Portugaly, Elon; Fromer, Menachem; Linial, Michal

    2008-07-01

    UPGMA (average linking) is probably the most popular algorithm for hierarchical data clustering, especially in computational biology. However, UPGMA requires the entire dissimilarity matrix in memory. Due to this prohibitive requirement, UPGMA is not scalable to very large datasets. We present a novel class of memory-constrained UPGMA (MC-UPGMA) algorithms. Given any practical memory size constraint, this framework guarantees the correct clustering solution without explicitly requiring all dissimilarities in memory. The algorithms are general and are applicable to any dataset. We present a data-dependent characterization of hardness and clustering efficiency. The presented concepts are applicable to any agglomerative clustering formulation. We apply our algorithm to the entire collection of protein sequences, to automatically build a comprehensive evolutionary-driven hierarchy of proteins from sequence alone. The newly created tree captures protein families better than state-of-the-art large-scale methods such as CluSTr, ProtoNet4 or single-linkage clustering. We demonstrate that leveraging the entire mass embodied in all sequence similarities allows to significantly improve on current protein family clusterings which are unable to directly tackle the sheer mass of this data. Furthermore, we argue that non-metric constraints are an inherent complexity of the sequence space and should not be overlooked. The robustness of UPGMA allows significant improvement, especially for multidomain proteins, and for large or divergent families. A comprehensive tree built from all UniProt sequence similarities, together with navigation and classification tools will be made available as part of the ProtoNet service. A C++ implementation of the algorithm is available on request.

  4. Hierarchical Control for Multiple DC-Microgrids Clusters

    DEFF Research Database (Denmark)

    Shafiee, Qobad; Dragicevic, Tomislav; Vasquez, Juan Carlos

    2014-01-01

    DC microgrids (MGs) have gained research interest during the recent years because of many potential advantages as compared to the ac system. To ensure reliable operation of a low-voltage dc MG as well as its intelligent operation with the other DC MGs, a hierarchical control is proposed in this p......DC microgrids (MGs) have gained research interest during the recent years because of many potential advantages as compared to the ac system. To ensure reliable operation of a low-voltage dc MG as well as its intelligent operation with the other DC MGs, a hierarchical control is proposed...

  5. The Case for a Hierarchical Cosmology

    Science.gov (United States)

    Vaucouleurs, G. de

    1970-01-01

    The development of modern theoretical cosmology is presented and some questionable assumptions of orthodox cosmology are pointed out. Suggests that recent observations indicate that hierarchical clustering is a basic factor in cosmology. The implications of hierarchical models of the universe are considered. Bibliography. (LC)

  6. Cluster analysis for applications

    CERN Document Server

    Anderberg, Michael R

    1973-01-01

    Cluster Analysis for Applications deals with methods and various applications of cluster analysis. Topics covered range from variables and scales to measures of association among variables and among data units. Conceptual problems in cluster analysis are discussed, along with hierarchical and non-hierarchical clustering methods. The necessary elements of data analysis, statistics, cluster analysis, and computer implementation are integrated vertically to cover the complete path from raw data to a finished analysis.Comprised of 10 chapters, this book begins with an introduction to the subject o

  7. Clustering analysis

    International Nuclear Information System (INIS)

    Romli

    1997-01-01

    Cluster analysis is the name of group of multivariate techniques whose principal purpose is to distinguish similar entities from the characteristics they process.To study this analysis, there are several algorithms that can be used. Therefore, this topic focuses to discuss the algorithms, such as, similarity measures, and hierarchical clustering which includes single linkage, complete linkage and average linkage method. also, non-hierarchical clustering method, which is popular name K -mean method ' will be discussed. Finally, this paper will be described the advantages and disadvantages of every methods

  8. Clustering, Hierarchical Organization, and the Topography of Abstract and Concrete Nouns

    Directory of Open Access Journals (Sweden)

    Joshua eTroche

    2014-04-01

    Full Text Available The empirical study of language has historically relied heavily upon concrete word stimuli. By definition, concrete words evoke salient perceptual associations that fit well within feature-based, sensorimotor models of word meaning. In contrast, many theorists argue that abstract words are disembodied in that their meaning is mediated through language. We investigated word meaning as distributed in multidimensional space using hierarchical cluster analysis. Participants (N=365 rated target words (n=400 English nouns across 12 cognitive dimensions (e.g., polarity, ease of teaching, emotional valence. Factor reduction revealed three latent factors, corresponding roughly to perceptual salience, affective association, and magnitude. We plotted the original 400 words for the three latent factors. Abstract and concrete words showed overlap in their topography but also differentiated themselves in semantic space. This topographic approach to word meaning offers a unique perspective to word concreteness.

  9. Clustering, hierarchical organization, and the topography of abstract and concrete nouns.

    Science.gov (United States)

    Troche, Joshua; Crutch, Sebastian; Reilly, Jamie

    2014-01-01

    The empirical study of language has historically relied heavily upon concrete word stimuli. By definition, concrete words evoke salient perceptual associations that fit well within feature-based, sensorimotor models of word meaning. In contrast, many theorists argue that abstract words are "disembodied" in that their meaning is mediated through language. We investigated word meaning as distributed in multidimensional space using hierarchical cluster analysis. Participants (N = 365) rated target words (n = 400 English nouns) across 12 cognitive dimensions (e.g., polarity, ease of teaching, emotional valence). Factor reduction revealed three latent factors, corresponding roughly to perceptual salience, affective association, and magnitude. We plotted the original 400 words for the three latent factors. Abstract and concrete words showed overlap in their topography but also differentiated themselves in semantic space. This topographic approach to word meaning offers a unique perspective to word concreteness.

  10. Unsupervised Classification of Surface Defects in Wire Rod Production Obtained by Eddy Current Sensors

    Directory of Open Access Journals (Sweden)

    Sergio Saludes-Rodil

    2015-04-01

    Full Text Available An unsupervised approach to classify surface defects in wire rod manufacturing is developed in this paper. The defects are extracted from an eddy current signal and classified using a clustering technique that uses the dynamic time warping distance as the dissimilarity measure. The new approach has been successfully tested using industrial data. It is shown that it outperforms other classification alternatives, such as the modified Fourier descriptors.

  11. Unsupervised Classification Using Immune Algorithm

    OpenAIRE

    Al-Muallim, M. T.; El-Kouatly, R.

    2012-01-01

    Unsupervised classification algorithm based on clonal selection principle named Unsupervised Clonal Selection Classification (UCSC) is proposed in this paper. The new proposed algorithm is data driven and self-adaptive, it adjusts its parameters to the data to make the classification operation as fast as possible. The performance of UCSC is evaluated by comparing it with the well known K-means algorithm using several artificial and real-life data sets. The experiments show that the proposed U...

  12. SDN‐Based Hierarchical Agglomerative Clustering Algorithm for Interference Mitigation in Ultra‐Dense Small Cell Networks

    Directory of Open Access Journals (Sweden)

    Guang Yang

    2018-04-01

    Full Text Available Ultra‐dense small cell networks (UD‐SCNs have been identified as a promising scheme for next‐generation wireless networks capable of meeting the ever‐increasing demand for higher transmission rates and better quality of service. However, UD‐SCNs will inevitably suffer from severe interference among the small cell base stations, which will lower their spectral efficiency. In this paper, we propose a software‐defined networking (SDN‐based hierarchical agglomerative clustering (SDN‐HAC framework, which leverages SDN to centrally control all sub‐channels in the network, and decides on cluster merging using a similarity criterion based on a suitability function. We evaluate the proposed algorithm through simulation. The obtained results show that the proposed algorithm performs well and improves system payoff by 18.19% and 436.34% when compared with the traditional network architecture algorithms and non‐cooperative scenarios, respectively.

  13. Defining functioning levels in patients with schizophrenia: A combination of a novel clustering method and brain SPECT analysis.

    Science.gov (United States)

    Catherine, Faget-Agius; Aurélie, Vincenti; Eric, Guedj; Pierre, Michel; Raphaëlle, Richieri; Marine, Alessandrini; Pascal, Auquier; Christophe, Lançon; Laurent, Boyer

    2017-12-30

    This study aims to define functioning levels of patients with schizophrenia by using a method of interpretable clustering based on a specific functioning scale, the Functional Remission Of General Schizophrenia (FROGS) scale, and to test their validity regarding clinical and neuroimaging characterization. In this observational study, patients with schizophrenia have been classified using a hierarchical top-down method called clustering using unsupervised binary trees (CUBT). Socio-demographic, clinical, and neuroimaging SPECT perfusion data were compared between the different clusters to ensure their clinical relevance. A total of 242 patients were analyzed. A four-group functioning level structure has been identified: 54 are classified as "minimal", 81 as "low", 64 as "moderate", and 43 as "high". The clustering shows satisfactory statistical properties, including reproducibility and discriminancy. The 4 clusters consistently differentiate patients. "High" functioning level patients reported significantly the lowest scores on the PANSS and the CDSS, and the highest scores on the GAF, the MARS and S-QoL 18. Functioning levels were significantly associated with cerebral perfusion of two relevant areas: the left inferior parietal cortex and the anterior cingulate. Our study provides relevant functioning levels in schizophrenia, and may enhance the use of functioning scale. Copyright © 2017 Elsevier B.V. All rights reserved.

  14. An Unsupervised Online Spike-Sorting Framework.

    Science.gov (United States)

    Knieling, Simeon; Sridharan, Kousik S; Belardinelli, Paolo; Naros, Georgios; Weiss, Daniel; Mormann, Florian; Gharabaghi, Alireza

    2016-08-01

    Extracellular neuronal microelectrode recordings can include action potentials from multiple neurons. To separate spikes from different neurons, they can be sorted according to their shape, a procedure referred to as spike-sorting. Several algorithms have been reported to solve this task. However, when clustering outcomes are unsatisfactory, most of them are difficult to adjust to achieve the desired results. We present an online spike-sorting framework that uses feature normalization and weighting to maximize the distinctiveness between different spike shapes. Furthermore, multiple criteria are applied to either facilitate or prevent cluster fusion, thereby enabling experimenters to fine-tune the sorting process. We compare our method to established unsupervised offline (Wave_Clus (WC)) and online (OSort (OS)) algorithms by examining their performance in sorting various test datasets using two different scoring systems (AMI and the Adamos metric). Furthermore, we evaluate sorting capabilities on intra-operative recordings using established quality metrics. Compared to WC and OS, our algorithm achieved comparable or higher scores on average and produced more convincing sorting results for intra-operative datasets. Thus, the presented framework is suitable for both online and offline analysis and could substantially improve the quality of microelectrode-based data evaluation for research and clinical application.

  15. Hierarchical Controlled Remote State Preparation by Using a Four-Qubit Cluster State

    Science.gov (United States)

    Ma, Peng-Cheng; Chen, Gui-Bin; Li, Xiao-Wei; Zhan, You-Bang

    2018-02-01

    We propose a scheme for hierarchical controlled remote preparation of an arbitrary single-qubit state via a four-qubit cluster state as the quantum channel. In this scheme, a sender wishes to help three agents to remotely prepare a quantum state, respectively. The three agents are divided into two grades, that is, an agent is in the upper grade and other two agents are in the lower grade. In this process of remote state preparation, the agent of the upper grade only needs the assistance of any one of the other two agents for recovering the sender's original state, while an agent of the lower grade needs the collaboration of all the other two agents. In other words, the agents of two grades have different authorities to reconstruct sender's original state.

  16. Hierarchical Controlled Remote State Preparation by Using a Four-Qubit Cluster State

    Science.gov (United States)

    Ma, Peng-Cheng; Chen, Gui-Bin; Li, Xiao-Wei; Zhan, You-Bang

    2018-06-01

    We propose a scheme for hierarchical controlled remote preparation of an arbitrary single-qubit state via a four-qubit cluster state as the quantum channel. In this scheme, a sender wishes to help three agents to remotely prepare a quantum state, respectively. The three agents are divided into two grades, that is, an agent is in the upper grade and other two agents are in the lower grade. In this process of remote state preparation, the agent of the upper grade only needs the assistance of any one of the other two agents for recovering the sender's original state, while an agent of the lower grade needs the collaboration of all the other two agents. In other words, the agents of two grades have different authorities to reconstruct sender's original state.

  17. Semi-Supervised Clustering for High-Dimensional and Sparse Features

    Science.gov (United States)

    Yan, Su

    2010-01-01

    Clustering is one of the most common data mining tasks, used frequently for data organization and analysis in various application domains. Traditional machine learning approaches to clustering are fully automated and unsupervised where class labels are unknown a priori. In real application domains, however, some "weak" form of side…

  18. Identifying influential individuals on intensive care units: using cluster analysis to explore culture.

    Science.gov (United States)

    Fong, Allan; Clark, Lindsey; Cheng, Tianyi; Franklin, Ella; Fernandez, Nicole; Ratwani, Raj; Parker, Sarah Henrickson

    2017-07-01

    The objective of this paper is to identify attribute patterns of influential individuals in intensive care units using unsupervised cluster analysis. Despite the acknowledgement that culture of an organisation is critical to improving patient safety, specific methods to shift culture have not been explicitly identified. A social network analysis survey was conducted and an unsupervised cluster analysis was used. A total of 100 surveys were gathered. Unsupervised cluster analysis was used to group individuals with similar dimensions highlighting three general genres of influencers: well-rounded, knowledge and relational. Culture is created locally by individual influencers. Cluster analysis is an effective way to identify common characteristics among members of an intensive care unit team that are noted as highly influential by their peers. To change culture, identifying and then integrating the influencers in intervention development and dissemination may create more sustainable and effective culture change. Additional studies are ongoing to test the effectiveness of utilising these influencers to disseminate patient safety interventions. This study offers an approach that can be helpful in both identifying and understanding influential team members and may be an important aspect of developing methods to change organisational culture. © 2017 John Wiley & Sons Ltd.

  19. Hydrothermal synthesis and photoluminescent properties of hierarchical GdPO4·H2O:Ln3+ (Ln3+ = Eu3+, Ce3+, Tb3+) flower-like clusters

    Science.gov (United States)

    Amurisana, Bao.; Zhiqiang, Song.; Haschaolu, O.; Yi, Chen; Tegus, O.

    2018-02-01

    3D hierarchical GdPO4·H2O:Ln3+ (Ln3+ = Eu3+, Ce3+, Tb3+) flower clusters were successfully prepared on glass slide substrate by a simple, economical hydrothermal process with the assistance of disodium ethylenediaminetetraacetic acid (Na2H2L, where L4- = (CH2COO)2N(CH2)2N(CH2COO)24-). In this process, Na2H2L was used as both a chelating agent and a structure-director. The hierarchical flower clusters have an average diameter of 7-12 μm and are composed of well-aligned microrods. The influence of the molar ratio of Na2H2L/Gd3+ and reaction time on the morphology was systematically studied. A possible crystal growth and formation mechanism of hierarchical flower clusters is proposed based on the evolution of morphology as a function of reaction time. The self-assembled GdPO4·H2O:Ln3+ superstructures exhibit strong orange-red (Eu3+, 5D0 → 7F1), green (Tb3+, 5D4 → 7F5) and near ultraviolet emissions (Ce3+, 5d → 7F5/2) under ultraviolet excitation, respectively. This study may provide a new channel for building hierarchically superstructued oxide micro/nanomaterials with optical and new properties.

  20. Micromechanics of hierarchical materials

    DEFF Research Database (Denmark)

    Mishnaevsky, Leon, Jr.

    2012-01-01

    A short overview of micromechanical models of hierarchical materials (hybrid composites, biomaterials, fractal materials, etc.) is given. Several examples of the modeling of strength and damage in hierarchical materials are summarized, among them, 3D FE model of hybrid composites...... with nanoengineered matrix, fiber bundle model of UD composites with hierarchically clustered fibers and 3D multilevel model of wood considered as a gradient, cellular material with layered composite cell walls. The main areas of research in micromechanics of hierarchical materials are identified, among them......, the investigations of the effects of load redistribution between reinforcing elements at different scale levels, of the possibilities to control different material properties and to ensure synergy of strengthening effects at different scale levels and using the nanoreinforcement effects. The main future directions...

  1. A Hybrid Supervised/Unsupervised Machine Learning Approach to Solar Flare Prediction

    Science.gov (United States)

    Benvenuto, Federico; Piana, Michele; Campi, Cristina; Massone, Anna Maria

    2018-01-01

    This paper introduces a novel method for flare forecasting, combining prediction accuracy with the ability to identify the most relevant predictive variables. This result is obtained by means of a two-step approach: first, a supervised regularization method for regression, namely, LASSO is applied, where a sparsity-enhancing penalty term allows the identification of the significance with which each data feature contributes to the prediction; then, an unsupervised fuzzy clustering technique for classification, namely, Fuzzy C-Means, is applied, where the regression outcome is partitioned through the minimization of a cost function and without focusing on the optimization of a specific skill score. This approach is therefore hybrid, since it combines supervised and unsupervised learning; realizes classification in an automatic, skill-score-independent way; and provides effective prediction performances even in the case of imbalanced data sets. Its prediction power is verified against NOAA Space Weather Prediction Center data, using as a test set, data in the range between 1996 August and 2010 December and as training set, data in the range between 1988 December and 1996 June. To validate the method, we computed several skill scores typically utilized in flare prediction and compared the values provided by the hybrid approach with the ones provided by several standard (non-hybrid) machine learning methods. The results showed that the hybrid approach performs classification better than all other supervised methods and with an effectiveness comparable to the one of clustering methods; but, in addition, it provides a reliable ranking of the weights with which the data properties contribute to the forecast.

  2. Unsupervised Learning (Clustering) of Odontocete Echolocation Clicks

    Science.gov (United States)

    2015-09-30

    develop methods for clustering of marine mammal echolocation clicks to learn about species assemblages where little or no prior knowledge exists about... Mexico or the Atlanic. 2 APPROACH Acoustic encounters with odontocetes are detected automatically and noise-corrected cepstral features...Estmation of Marine Mammals Using Passive Acoustic Monitoring (DCLDE). KL divergence maps were created for all known species, but the sperm whale

  3. Complex scenes and situations visualization in hierarchical learning algorithm with dynamic 3D NeoAxis engine

    Science.gov (United States)

    Graham, James; Ternovskiy, Igor V.

    2013-06-01

    We applied a two stage unsupervised hierarchical learning system to model complex dynamic surveillance and cyber space monitoring systems using a non-commercial version of the NeoAxis visualization software. The hierarchical scene learning and recognition approach is based on hierarchical expectation maximization, and was linked to a 3D graphics engine for validation of learning and classification results and understanding the human - autonomous system relationship. Scene recognition is performed by taking synthetically generated data and feeding it to a dynamic logic algorithm. The algorithm performs hierarchical recognition of the scene by first examining the features of the objects to determine which objects are present, and then determines the scene based on the objects present. This paper presents a framework within which low level data linked to higher-level visualization can provide support to a human operator and be evaluated in a detailed and systematic way.

  4. Discovering hierarchical structure in normal relational data

    DEFF Research Database (Denmark)

    Schmidt, Mikkel Nørgaard; Herlau, Tue; Mørup, Morten

    2014-01-01

    -parametric generative model for hierarchical clustering of similarity based on multifurcating Gibbs fragmentation trees. This allows us to infer and display the posterior distribution of hierarchical structures that comply with the data. We demonstrate the utility of our method on synthetic data and data of functional...

  5. DATA CLASSIFICATION WITH NEURAL CLASSIFIER USING RADIAL BASIS FUNCTION WITH DATA REDUCTION USING HIERARCHICAL CLUSTERING

    Directory of Open Access Journals (Sweden)

    M. Safish Mary

    2012-04-01

    Full Text Available Classification of large amount of data is a time consuming process but crucial for analysis and decision making. Radial Basis Function networks are widely used for classification and regression analysis. In this paper, we have studied the performance of RBF neural networks to classify the sales of cars based on the demand, using kernel density estimation algorithm which produces classification accuracy comparable to data classification accuracy provided by support vector machines. In this paper, we have proposed a new instance based data selection method where redundant instances are removed with help of a threshold thus improving the time complexity with improved classification accuracy. The instance based selection of the data set will help reduce the number of clusters formed thereby reduces the number of centers considered for building the RBF network. Further the efficiency of the training is improved by applying a hierarchical clustering technique to reduce the number of clusters formed at every step. The paper explains the algorithm used for classification and for conditioning the data. It also explains the complexities involved in classification of sales data for analysis and decision-making.

  6. Examination of Clustering in Eutectic Microstrcture

    Directory of Open Access Journals (Sweden)

    Bortnyik K.

    2017-06-01

    Full Text Available The eutectic microstructures are complex microstructures and a hard work to describe it with few numbers. The eutectics builds up eutectic cells. In the cells the phases are clustered. With the development of big databases the data mining also develops, and produces a lot of method to handling the large datasets, and earns information from the sets. One typical method is the clustering, which finds the groups in the datasets. In this article a partitioning and a hierarchical clustering is applied to eutectic structures to find the clusters. In the case of AlMn alloy the K-means algorithm work well, and find the eutectic cells. In the case of ductile cast iron the hierarchical clustering works better. With the combination of the partitioning and hierarchical clustering with the image transformation, an effective method is developed for clustering the objects in the microstructures.

  7. Unsupervised Image Segmentation

    Czech Academy of Sciences Publication Activity Database

    Haindl, Michal; Mikeš, Stanislav

    2014-01-01

    Roč. 36, č. 4 (2014), s. 23-23 R&D Projects: GA ČR(CZ) GA14-10911S Institutional support: RVO:67985556 Keywords : unsupervised image segmentation Subject RIV: BD - Theory of Information http://library.utia.cas.cz/separaty/2014/RO/haindl-0434412.pdf

  8. Semi-supervised clustering methods.

    Science.gov (United States)

    Bair, Eric

    2013-01-01

    Cluster analysis methods seek to partition a data set into homogeneous subgroups. It is useful in a wide variety of applications, including document processing and modern genetics. Conventional clustering methods are unsupervised, meaning that there is no outcome variable nor is anything known about the relationship between the observations in the data set. In many situations, however, information about the clusters is available in addition to the values of the features. For example, the cluster labels of some observations may be known, or certain observations may be known to belong to the same cluster. In other cases, one may wish to identify clusters that are associated with a particular outcome variable. This review describes several clustering algorithms (known as "semi-supervised clustering" methods) that can be applied in these situations. The majority of these methods are modifications of the popular k-means clustering method, and several of them will be described in detail. A brief description of some other semi-supervised clustering algorithms is also provided.

  9. A new hybrid imperialist competitive algorithm on data clustering

    Indian Academy of Sciences (India)

    Modified imperialist competitive algorithm; simulated annealing; ... Clustering is one of the unsupervised learning branches where a set of patterns, usually vectors ..... machine classification is based on design, operation, and/or purpose.

  10. A Negative Selection Algorithm Based on Hierarchical Clustering of Self Set and its Application in Anomaly Detection

    Directory of Open Access Journals (Sweden)

    Wen Chen

    2011-08-01

    Full Text Available A negative selection algorithm based on the hierarchical clustering of self set HC-RNSA is introduced in this paper. Several strategies are applied to improve the algorithm performance. First, the self data set is replaced by the self cluster centers to compare with the detector candidates in each cluster level. As the number of self clusters is much less than the self set size, the detector generation efficiency is improved. Second, during the detector generation process, the detector candidates are restricted to the lower coverage space to reduce detector redundancy. In the article, the problem that the distances between antigens coverage to a constant value in the high dimensional space is analyzed, accordingly the Principle Component Analysis (PCA method is used to reduce the data dimension, and the fractional distance function is employed to enhance the distinctiveness between the self and non-self antigens. The detector generation procedure is terminated when the expected non-self coverage is reached. The theory analysis and experimental results demonstrate that the detection rate of HC-RNSA is higher than that of the traditional negative selection algorithms while the false alarm rate and time cost are reduced.

  11. Unsupervised action classification using space-time link analysis

    DEFF Research Database (Denmark)

    Liu, Haowei; Feris, Rogerio; Krüger, Volker

    2010-01-01

    In this paper we address the problem of unsupervised discovery of action classes in video data. Different from all existing methods thus far proposed for this task, we present a space-time link analysis approach which matches the performance of traditional unsupervised action categorization metho...

  12. An improved clustering algorithm based on reverse learning in intelligent transportation

    Science.gov (United States)

    Qiu, Guoqing; Kou, Qianqian; Niu, Ting

    2017-05-01

    With the development of artificial intelligence and data mining technology, big data has gradually entered people's field of vision. In the process of dealing with large data, clustering is an important processing method. By introducing the reverse learning method in the clustering process of PAM clustering algorithm, to further improve the limitations of one-time clustering in unsupervised clustering learning, and increase the diversity of clustering clusters, so as to improve the quality of clustering. The algorithm analysis and experimental results show that the algorithm is feasible.

  13. Unsupervised EEG analysis for automated epileptic seizure detection

    Science.gov (United States)

    Birjandtalab, Javad; Pouyan, Maziyar Baran; Nourani, Mehrdad

    2016-07-01

    Epilepsy is a neurological disorder which can, if not controlled, potentially cause unexpected death. It is extremely crucial to have accurate automatic pattern recognition and data mining techniques to detect the onset of seizures and inform care-givers to help the patients. EEG signals are the preferred biosignals for diagnosis of epileptic patients. Most of the existing pattern recognition techniques used in EEG analysis leverage the notion of supervised machine learning algorithms. Since seizure data are heavily under-represented, such techniques are not always practical particularly when the labeled data is not sufficiently available or when disease progression is rapid and the corresponding EEG footprint pattern will not be robust. Furthermore, EEG pattern change is highly individual dependent and requires experienced specialists to annotate the seizure and non-seizure events. In this work, we present an unsupervised technique to discriminate seizures and non-seizures events. We employ power spectral density of EEG signals in different frequency bands that are informative features to accurately cluster seizure and non-seizure events. The experimental results tried so far indicate achieving more than 90% accuracy in clustering seizure and non-seizure events without having any prior knowledge on patient's history.

  14. Do COPD subtypes really exist? COPD heterogeneity and clustering in 10 independent cohorts

    NARCIS (Netherlands)

    Castaldi, Peter J; Benet, Marta; Petersen, Hans; Rafaels, Nicholas; Finigan, James; Paoletti, Matteo; Marike Boezen, H; Vonk, Judith M; Bowler, Russell; Pistolesi, Massimo; Puhan, Milo A; Anto, Josep; Wauters, Els; Lambrechts, Diether; Janssens, Wim; Bigazzi, Francesca; Camiciottoli, Gianna; Cho, Michael H; Hersh, Craig P; Barnes, Kathleen; Rennard, Stephen; Boorgula, Meher Preethi; Dy, Jennifer; Hansel, Nadia N; Crapo, James D; Tesfaigzi, Yohannes; Agusti, Alvar; Silverman, Edwin K; Garcia-Aymerich, Judith

    Background COPD is a heterogeneous disease, but there is little consensus on specific definitions for COPD subtypes. Unsupervised clustering offers the promise of 'unbiased' data-driven assessment of COPD heterogeneity. Multiple groups have identified COPD subtypes using cluster analysis, but there

  15. Semi-Supervised Generation with Cluster-aware Generative Models

    DEFF Research Database (Denmark)

    Maaløe, Lars; Fraccaro, Marco; Winther, Ole

    2017-01-01

    Deep generative models trained with large amounts of unlabelled data have proven to be powerful within the domain of unsupervised learning. Many real life data sets contain a small amount of labelled data points, that are typically disregarded when training generative models. We propose the Clust...... a log-likelihood of −79.38 nats on permutation invariant MNIST, while also achieving competitive semi-supervised classification accuracies. The model can also be trained fully unsupervised, and still improve the log-likelihood performance with respect to related methods.......Deep generative models trained with large amounts of unlabelled data have proven to be powerful within the domain of unsupervised learning. Many real life data sets contain a small amount of labelled data points, that are typically disregarded when training generative models. We propose the Cluster...

  16. A new web-based system for unsupervised classification of satellite images from the Google Maps engine

    Science.gov (United States)

    Ferrán, Ángel; Bernabé, Sergio; García-Rodríguez, Pablo; Plaza, Antonio

    2012-10-01

    In this paper, we develop a new web-based system for unsupervised classification of satellite images available from the Google Maps engine. The system has been developed using the Google Maps API and incorporates functionalities such as unsupervised classification of image portions selected by the user (at the desired zoom level). For this purpose, we use a processing chain made up of the well-known ISODATA and k-means algorithms, followed by spatial post-processing based on majority voting. The system is currently hosted on a high performance server which performs the execution of classification algorithms and returns the obtained classification results in a very efficient way. The previous functionalities are necessary to use efficient techniques for the classification of images and the incorporation of content-based image retrieval (CBIR). Several experimental validation types of the classification results with the proposed system are performed by comparing the classification accuracy of the proposed chain by means of techniques available in the well-known Environment for Visualizing Images (ENVI) software package. The server has access to a cluster of commodity graphics processing units (GPUs), hence in future work we plan to perform the processing in parallel by taking advantage of the cluster.

  17. Unsupervised Performance Evaluation of Image Segmentation

    Directory of Open Access Journals (Sweden)

    Chabrier Sebastien

    2006-01-01

    Full Text Available We present in this paper a study of unsupervised evaluation criteria that enable the quantification of the quality of an image segmentation result. These evaluation criteria compute some statistics for each region or class in a segmentation result. Such an evaluation criterion can be useful for different applications: the comparison of segmentation results, the automatic choice of the best fitted parameters of a segmentation method for a given image, or the definition of new segmentation methods by optimization. We first present the state of art of unsupervised evaluation, and then, we compare six unsupervised evaluation criteria. For this comparative study, we use a database composed of 8400 synthetic gray-level images segmented in four different ways. Vinet's measure (correct classification rate is used as an objective criterion to compare the behavior of the different criteria. Finally, we present the experimental results on the segmentation evaluation of a few gray-level natural images.

  18. Unsupervised Learning in an Ensemble of Spiking Neural Networks Mediated by ITDP.

    Directory of Open Access Journals (Sweden)

    Yoonsik Shim

    2016-10-01

    Full Text Available We propose a biologically plausible architecture for unsupervised ensemble learning in a population of spiking neural network classifiers. A mixture of experts type organisation is shown to be effective, with the individual classifier outputs combined via a gating network whose operation is driven by input timing dependent plasticity (ITDP. The ITDP gating mechanism is based on recent experimental findings. An abstract, analytically tractable model of the ITDP driven ensemble architecture is derived from a logical model based on the probabilities of neural firing events. A detailed analysis of this model provides insights that allow it to be extended into a full, biologically plausible, computational implementation of the architecture which is demonstrated on a visual classification task. The extended model makes use of a style of spiking network, first introduced as a model of cortical microcircuits, that is capable of Bayesian inference, effectively performing expectation maximization. The unsupervised ensemble learning mechanism, based around such spiking expectation maximization (SEM networks whose combined outputs are mediated by ITDP, is shown to perform the visual classification task well and to generalize to unseen data. The combined ensemble performance is significantly better than that of the individual classifiers, validating the ensemble architecture and learning mechanisms. The properties of the full model are analysed in the light of extensive experiments with the classification task, including an investigation into the influence of different input feature selection schemes and a comparison with a hierarchical STDP based ensemble architecture.

  19. Unsupervised Learning in an Ensemble of Spiking Neural Networks Mediated by ITDP.

    Science.gov (United States)

    Shim, Yoonsik; Philippides, Andrew; Staras, Kevin; Husbands, Phil

    2016-10-01

    We propose a biologically plausible architecture for unsupervised ensemble learning in a population of spiking neural network classifiers. A mixture of experts type organisation is shown to be effective, with the individual classifier outputs combined via a gating network whose operation is driven by input timing dependent plasticity (ITDP). The ITDP gating mechanism is based on recent experimental findings. An abstract, analytically tractable model of the ITDP driven ensemble architecture is derived from a logical model based on the probabilities of neural firing events. A detailed analysis of this model provides insights that allow it to be extended into a full, biologically plausible, computational implementation of the architecture which is demonstrated on a visual classification task. The extended model makes use of a style of spiking network, first introduced as a model of cortical microcircuits, that is capable of Bayesian inference, effectively performing expectation maximization. The unsupervised ensemble learning mechanism, based around such spiking expectation maximization (SEM) networks whose combined outputs are mediated by ITDP, is shown to perform the visual classification task well and to generalize to unseen data. The combined ensemble performance is significantly better than that of the individual classifiers, validating the ensemble architecture and learning mechanisms. The properties of the full model are analysed in the light of extensive experiments with the classification task, including an investigation into the influence of different input feature selection schemes and a comparison with a hierarchical STDP based ensemble architecture.

  20. Investigating the effects of climate variations on bacillary dysentery incidence in northeast China using ridge regression and hierarchical cluster analysis

    Directory of Open Access Journals (Sweden)

    Guo Junqiao

    2008-09-01

    Full Text Available Abstract Background The effects of climate variations on bacillary dysentery incidence have gained more recent concern. However, the multi-collinearity among meteorological factors affects the accuracy of correlation with bacillary dysentery incidence. Methods As a remedy, a modified method to combine ridge regression and hierarchical cluster analysis was proposed for investigating the effects of climate variations on bacillary dysentery incidence in northeast China. Results All weather indicators, temperatures, precipitation, evaporation and relative humidity have shown positive correlation with the monthly incidence of bacillary dysentery, while air pressure had a negative correlation with the incidence. Ridge regression and hierarchical cluster analysis showed that during 1987–1996, relative humidity, temperatures and air pressure affected the transmission of the bacillary dysentery. During this period, all meteorological factors were divided into three categories. Relative humidity and precipitation belonged to one class, temperature indexes and evaporation belonged to another class, and air pressure was the third class. Conclusion Meteorological factors have affected the transmission of bacillary dysentery in northeast China. Bacillary dysentery prevention and control would benefit from by giving more consideration to local climate variations.

  1. Characterizing Interference in Radio Astronomy Observations through Active and Unsupervised Learning

    Science.gov (United States)

    Doran, G.

    2013-01-01

    In the process of observing signals from astronomical sources, radio astronomers must mitigate the effects of manmade radio sources such as cell phones, satellites, aircraft, and observatory equipment. Radio frequency interference (RFI) often occurs as short bursts (active learning approach in which an astronomer labels events that are most confusing to a classifier, minimizing the human effort required for classification. We also explore the use of unsupervised clustering techniques, which automatically group events into classes without user input. We apply these techniques to data from the Parkes Multibeam Pulsar Survey to characterize several million detected RFI events from over a thousand hours of observation.

  2. Genome-wide decoding of hierarchical modular structure of transcriptional regulation by cis-element and expression clustering.

    Science.gov (United States)

    Leyfer, Dmitriy; Weng, Zhiping

    2005-09-01

    A holistic approach to the study of cellular processes is identifying both gene-expression changes and regulatory elements promoting such changes. Cellular regulatory processes can be viewed as transcriptional modules (TMs), groups of coexpressed genes regulated by groups of transcription factors (TFs). We set out to devise a method that would identify TMs while avoiding arbitrary thresholds on TM sizes and number. Assuming that gene expression is determined by TFs that bind to the gene's promoter, clustering of genes based on TF binding sites (cis-elements) should create gene groups similar to those obtained by gene expression clustering. Intersections between the expression and cis-element-based gene clusters reveal TMs. Statistical significance assigned to each TM allows identification of regulatory units of any size. Our method correctly identifies the number and sizes of TMs on simulated datasets. We demonstrate that yeast experimental TMs are biologically relevant by comparing them with MIPS and GO categories. Our modules are in statistically significant agreement with TMs from other research groups. This work suggests that there is no preferential division of biological processes into regulatory units; each degree of partitioning exhibits a slice of biological network revealing hierarchical modular organization of transcriptional regulation.

  3. A Performance-Prediction Model for PIC Applications on Clusters of Symmetric MultiProcessors: Validation with Hierarchical HPF+OpenMP Implementation

    Directory of Open Access Journals (Sweden)

    Sergio Briguglio

    2003-01-01

    Full Text Available A performance-prediction model is presented, which describes different hierarchical workload decomposition strategies for particle in cell (PIC codes on Clusters of Symmetric MultiProcessors. The devised workload decomposition is hierarchically structured: a higher-level decomposition among the computational nodes, and a lower-level one among the processors of each computational node. Several decomposition strategies are evaluated by means of the prediction model, with respect to the memory occupancy, the parallelization efficiency and the required programming effort. Such strategies have been implemented by integrating the high-level languages High Performance Fortran (at the inter-node stage and OpenMP (at the intra-node one. The details of these implementations are presented, and the experimental values of parallelization efficiency are compared with the predicted results.

  4. "Analyzing the Longitudinal K-12 Grading Histories of Entire Cohorts of Students: Grades, Data Driven Decision Making, Dropping out and Hierarchical Cluster Analysis"

    Directory of Open Access Journals (Sweden)

    Alex J. Bowers

    2010-05-01

    Full Text Available School personnel currently lack an effective method to pattern and visually interpret disaggregated achievement data collected on students as a means to help inform decision making. This study, through the examination of longitudinal K-12 teacher assigned grading histories for entire cohorts of students from a school district (n=188, demonstrates a novel application of hierarchical cluster analysis and pattern visualization in which all data points collected on every student in a cohort can be patterned, visualized and interpreted to aid in data driven decision making by teachers and administrators. Additionally, as a proof-of-concept study, overall schooling outcomes, such as student dropout or taking a college entrance exam, are identified from the data patterns and compared to past methods of dropout identification as one example of the usefulness of the method. Hierarchical cluster analysis correctly identified over 80% of the students who dropped out using the entire student grade history patterns from either K-12 or K-8.

  5. Multispectral and Panchromatic used Enhancement Resolution and Study Effective Enhancement on Supervised and Unsupervised Classification Land – Cover

    Science.gov (United States)

    Salman, S. S.; Abbas, W. A.

    2018-05-01

    The goal of the study is to support analysis Enhancement of Resolution and study effect on classification methods on bands spectral information of specific and quantitative approaches. In this study introduce a method to enhancement resolution Landsat 8 of combining the bands spectral of 30 meters resolution with panchromatic band 8 of 15 meters resolution, because of importance multispectral imagery to extracting land - cover. Classification methods used in this study to classify several lands -covers recorded from OLI- 8 imagery. Two methods of Data mining can be classified as either supervised or unsupervised. In supervised methods, there is a particular predefined target, that means the algorithm learn which values of the target are associated with which values of the predictor sample. K-nearest neighbors and maximum likelihood algorithms examine in this work as supervised methods. In other hand, no sample identified as target in unsupervised methods, the algorithm of data extraction searches for structure and patterns between all the variables, represented by Fuzzy C-mean clustering method as one of the unsupervised methods, NDVI vegetation index used to compare the results of classification method, the percent of dense vegetation in maximum likelihood method give a best results.

  6. Detecting Transitions in Manual Tasks from Wearables: An Unsupervised Labeling Approach

    Directory of Open Access Journals (Sweden)

    Sebastian Böttcher

    2018-03-01

    Full Text Available Authoring protocols for manual tasks such as following recipes, manufacturing processes or laboratory experiments requires significant effort. This paper presents a system that estimates individual procedure transitions from the user’s physical movement and gestures recorded with inertial motion sensors. Combined with egocentric or external video recordings, this facilitates efficient review and annotation of video databases. We investigate different clustering algorithms on wearable inertial sensor data recorded on par with video data, to automatically create transition marks between task steps. The goal is to match these marks to the transitions given in a description of the workflow, thus creating navigation cues to browse video repositories of manual work. To evaluate the performance of unsupervised algorithms, the automatically-generated marks are compared to human expert-created labels on two publicly-available datasets. Additionally, we tested the approach on a novel dataset in a manufacturing lab environment, describing an existing sequential manufacturing process. The results from selected clustering methods are also compared to some supervised methods.

  7. Automatic Clustering Using FSDE-Forced Strategy Differential Evolution

    Science.gov (United States)

    Yasid, A.

    2018-01-01

    Clustering analysis is important in datamining for unsupervised data, cause no adequate prior knowledge. One of the important tasks is defining the number of clusters without user involvement that is known as automatic clustering. This study intends on acquiring cluster number automatically utilizing forced strategy differential evolution (AC-FSDE). Two mutation parameters, namely: constant parameter and variable parameter are employed to boost differential evolution performance. Four well-known benchmark datasets were used to evaluate the algorithm. Moreover, the result is compared with other state of the art automatic clustering methods. The experiment results evidence that AC-FSDE is better or competitive with other existing automatic clustering algorithm.

  8. Automated three-dimensional morphology-based clustering of human erythrocytes with regular shapes: stomatocytes, discocytes, and echinocytes

    Science.gov (United States)

    Ahmadzadeh, Ezat; Jaferzadeh, Keyvan; Lee, Jieun; Moon, Inkyu

    2017-07-01

    We present unsupervised clustering methods for automatic grouping of human red blood cells (RBCs) extracted from RBC quantitative phase images obtained by digital holographic microscopy into three RBC clusters with regular shapes, including biconcave, stomatocyte, and sphero-echinocyte. We select some good features related to the RBC profile and morphology, such as RBC average thickness, sphericity coefficient, and mean corpuscular volume, and clustering methods, including density-based spatial clustering applications with noise, k-medoids, and k-means, are applied to the set of morphological features. The clustering results of RBCs using a set of three-dimensional features are compared against a set of two-dimensional features. Our experimental results indicate that by utilizing the introduced set of features, two groups of biconcave RBCs and old RBCs (suffering from the sphero-echinocyte process) can be perfectly clustered. In addition, by increasing the number of clusters, the three RBC types can be effectively clustered in an automated unsupervised manner with high accuracy. The performance evaluation of the clustering techniques reveals that they can assist hematologists in further diagnosis.

  9. Topology of the correlation networks among major currencies using hierarchical structure methods

    Science.gov (United States)

    Keskin, Mustafa; Deviren, Bayram; Kocakaplan, Yusuf

    2011-02-01

    We studied the topology of correlation networks among 34 major currencies using the concept of a minimal spanning tree and hierarchical tree for the full years of 2007-2008 when major economic turbulence occurred. We used the USD (US Dollar) and the TL (Turkish Lira) as numeraires in which the USD was the major currency and the TL was the minor currency. We derived a hierarchical organization and constructed minimal spanning trees (MSTs) and hierarchical trees (HTs) for the full years of 2007, 2008 and for the 2007-2008 period. We performed a technique to associate a value of reliability to the links of MSTs and HTs by using bootstrap replicas of data. We also used the average linkage cluster analysis for obtaining the hierarchical trees in the case of the TL as the numeraire. These trees are useful tools for understanding and detecting the global structure, taxonomy and hierarchy in financial data. We illustrated how the minimal spanning trees and their related hierarchical trees developed over a period of time. From these trees we identified different clusters of currencies according to their proximity and economic ties. The clustered structure of the currencies and the key currency in each cluster were obtained and we found that the clusters matched nicely with the geographical regions of corresponding countries in the world such as Asia or Europe. As expected the key currencies were generally those showing major economic activity.

  10. Unsupervised Condition Change Detection In Large Diesel Engines

    DEFF Research Database (Denmark)

    Pontoppidan, Niels Henrik; Larsen, Jan

    2003-01-01

    This paper presents a new method for unsupervised change detection which combines independent component modeling and probabilistic outlier etection. The method further provides a compact data representation, which is amenable to interpretation, i.e., the detected condition changes can be investig...... be investigated further. The method is successfully applied to unsupervised condition change detection in large diesel engines from acoustical emission sensor signal and compared to more classical techniques based on principal component analysis and Gaussian mixture models.......This paper presents a new method for unsupervised change detection which combines independent component modeling and probabilistic outlier etection. The method further provides a compact data representation, which is amenable to interpretation, i.e., the detected condition changes can...

  11. Hierarchical Sets: Analyzing Pangenome Structure through Scalable Set Visualizations

    DEFF Research Database (Denmark)

    Pedersen, Thomas Lin

    2017-01-01

    of hierarchical sets by applying it to a pangenome based on 113 Escherichia and Shigella genomes and find it provides a powerful addition to pangenome analysis. The described clustering algorithm and visualizations are implemented in the hierarchicalSets R package available from CRAN (https...

  12. A Comparison of Methods for Player Clustering via Behavioral Telemetry

    DEFF Research Database (Denmark)

    Drachen, Anders; Thurau, C.; Sifa, R.

    2013-01-01

    patterns in the behavioral data, and developing profiles that are actionable to game developers. There are numerous methods for unsupervised clustering of user behavior, e.g. k-means/c-means, Nonnegative Matrix Factorization, or Principal Component Analysis. Although all yield behavior categorizations......, interpretation of the resulting categories in terms of actual play behavior can be difficult if not impossible. In this paper, a range of unsupervised techniques are applied together with Archetypal Analysis to develop behavioral clusters from playtime data of 70,014 World of Warcraft players, covering a five......The analysis of user behavior in digital games has been aided by the introduction of user telemetry in game development, which provides unprecedented access to quantitative data on user behavior from the installed game clients of the entire population of players. Player behavior telemetry datasets...

  13. Modeling language and cognition with deep unsupervised learning: a tutorial overview.

    Science.gov (United States)

    Zorzi, Marco; Testolin, Alberto; Stoianov, Ivilin P

    2013-01-01

    Deep unsupervised learning in stochastic recurrent neural networks with many layers of hidden units is a recent breakthrough in neural computation research. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. In this article we discuss the theoretical foundations of this approach and we review key issues related to training, testing and analysis of deep networks for modeling language and cognitive processing. The classic letter and word perception problem of McClelland and Rumelhart (1981) is used as a tutorial example to illustrate how structured and abstract representations may emerge from deep generative learning. We argue that the focus on deep architectures and generative (rather than discriminative) learning represents a crucial step forward for the connectionist modeling enterprise, because it offers a more plausible model of cortical learning as well as a way to bridge the gap between emergentist connectionist models and structured Bayesian models of cognition.

  14. Modeling Language and Cognition with Deep Unsupervised Learning:A Tutorial Overview

    Directory of Open Access Journals (Sweden)

    Marco eZorzi

    2013-08-01

    Full Text Available Deep unsupervised learning in stochastic recurrent neural networks with many layers of hidden units is a recent breakthrough in neural computation research. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. In this article we discuss the theoretical foundations of this approach and we review key issues related to training, testing and analysis of deep networks for modeling language and cognitive processing. The classic letter and word perception problem of McClelland and Rumelhart (1981 is used as a tutorial example to illustrate how structured and abstract representations may emerge from deep generative learning. We argue that the focus on deep architectures and generative (rather than discriminative learning represents a crucial step forward for the connectionist modeling enterprise, because it offers a more plausible model of cortical learning as well as way to bridge the gap between emergentist connectionist models and structured Bayesian models of cognition.

  15. Modeling language and cognition with deep unsupervised learning: a tutorial overview

    Science.gov (United States)

    Zorzi, Marco; Testolin, Alberto; Stoianov, Ivilin P.

    2013-01-01

    Deep unsupervised learning in stochastic recurrent neural networks with many layers of hidden units is a recent breakthrough in neural computation research. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. In this article we discuss the theoretical foundations of this approach and we review key issues related to training, testing and analysis of deep networks for modeling language and cognitive processing. The classic letter and word perception problem of McClelland and Rumelhart (1981) is used as a tutorial example to illustrate how structured and abstract representations may emerge from deep generative learning. We argue that the focus on deep architectures and generative (rather than discriminative) learning represents a crucial step forward for the connectionist modeling enterprise, because it offers a more plausible model of cortical learning as well as a way to bridge the gap between emergentist connectionist models and structured Bayesian models of cognition. PMID:23970869

  16. Weighted Clustering

    DEFF Research Database (Denmark)

    Ackerman, Margareta; Ben-David, Shai; Branzei, Simina

    2012-01-01

    We investigate a natural generalization of the classical clustering problem, considering clustering tasks in which different instances may have different weights.We conduct the first extensive theoretical analysis on the influence of weighted data on standard clustering algorithms in both...... the partitional and hierarchical settings, characterizing the conditions under which algorithms react to weights. Extending a recent framework for clustering algorithm selection, we propose intuitive properties that would allow users to choose between clustering algorithms in the weighted setting and classify...

  17. Semi-supervised clustering methods

    Science.gov (United States)

    Bair, Eric

    2013-01-01

    Cluster analysis methods seek to partition a data set into homogeneous subgroups. It is useful in a wide variety of applications, including document processing and modern genetics. Conventional clustering methods are unsupervised, meaning that there is no outcome variable nor is anything known about the relationship between the observations in the data set. In many situations, however, information about the clusters is available in addition to the values of the features. For example, the cluster labels of some observations may be known, or certain observations may be known to belong to the same cluster. In other cases, one may wish to identify clusters that are associated with a particular outcome variable. This review describes several clustering algorithms (known as “semi-supervised clustering” methods) that can be applied in these situations. The majority of these methods are modifications of the popular k-means clustering method, and several of them will be described in detail. A brief description of some other semi-supervised clustering algorithms is also provided. PMID:24729830

  18. Generation of brain pseudo-CTs using an undersampled, single-acquisition UTE-mDixon pulse sequence and unsupervised clustering

    International Nuclear Information System (INIS)

    Su, Kuan-Hao; Hu, Lingzhi; Traughber, Melanie; Stehning, Christian; Helle, Michael; Qian, Pengjiang; Thompson, Cheryl L.; Pereira, Gisele C.; Traughber, Bryan J.; Jordan, David W.; Herrmann, Karin A.; Muzic, Raymond F.

    2015-01-01

    Purpose: MR-based pseudo-CT has an important role in MR-based radiation therapy planning and PET attenuation correction. The purpose of this study is to establish a clinically feasible approach, including image acquisition, correction, and CT formation, for pseudo-CT generation of the brain using a single-acquisition, undersampled ultrashort echo time (UTE)-mDixon pulse sequence. Methods: Nine patients were recruited for this study. For each patient, a 190-s, undersampled, single acquisition UTE-mDixon sequence of the brain was acquired (TE = 0.1, 1.5, and 2.8 ms). A novel method of retrospective trajectory correction of the free induction decay (FID) signal was performed based on point-spread functions of three external MR markers. Two-point Dixon images were reconstructed using the first and second echo data (TE = 1.5 and 2.8 ms). R2 ∗ images (1/T2 ∗ ) were then estimated and were used to provide bone information. Three image features, i.e., Dixon-fat, Dixon-water, and R2 ∗ , were used for unsupervised clustering. Five tissue clusters, i.e., air, brain, fat, fluid, and bone, were estimated using the fuzzy c-means (FCM) algorithm. A two-step, automatic tissue-assignment approach was proposed and designed according to the prior information of the given feature space. Pseudo-CTs were generated by a voxelwise linear combination of the membership functions of the FCM. A low-dose CT was acquired for each patient and was used as the gold standard for comparison. Results: The contrast and sharpness of the FID images were improved after trajectory correction was applied. The mean of the estimated trajectory delay was 0.774 μs (max: 1.350 μs; min: 0.180 μs). The FCM-estimated centroids of different tissue types showed a distinguishable pattern for different tissues, and significant differences were found between the centroid locations of different tissue types. Pseudo-CT can provide additional skull detail and has low bias and absolute error of estimated CT

  19. Hierarchical Neural Regression Models for Customer Churn Prediction

    Directory of Open Access Journals (Sweden)

    Golshan Mohammadi

    2013-01-01

    Full Text Available As customers are the main assets of each industry, customer churn prediction is becoming a major task for companies to remain in competition with competitors. In the literature, the better applicability and efficiency of hierarchical data mining techniques has been reported. This paper considers three hierarchical models by combining four different data mining techniques for churn prediction, which are backpropagation artificial neural networks (ANN, self-organizing maps (SOM, alpha-cut fuzzy c-means (α-FCM, and Cox proportional hazards regression model. The hierarchical models are ANN + ANN + Cox, SOM + ANN + Cox, and α-FCM + ANN + Cox. In particular, the first component of the models aims to cluster data in two churner and nonchurner groups and also filter out unrepresentative data or outliers. Then, the clustered data as the outputs are used to assign customers to churner and nonchurner groups by the second technique. Finally, the correctly classified data are used to create Cox proportional hazards model. To evaluate the performance of the hierarchical models, an Iranian mobile dataset is considered. The experimental results show that the hierarchical models outperform the single Cox regression baseline model in terms of prediction accuracy, Types I and II errors, RMSE, and MAD metrics. In addition, the α-FCM + ANN + Cox model significantly performs better than the two other hierarchical models.

  20. Automatic microseismic event picking via unsupervised machine learning

    Science.gov (United States)

    Chen, Yangkang

    2018-01-01

    Effective and efficient arrival picking plays an important role in microseismic and earthquake data processing and imaging. Widely used short-term-average long-term-average ratio (STA/LTA) based arrival picking algorithms suffer from the sensitivity to moderate-to-strong random ambient noise. To make the state-of-the-art arrival picking approaches effective, microseismic data need to be first pre-processed, for example, removing sufficient amount of noise, and second analysed by arrival pickers. To conquer the noise issue in arrival picking for weak microseismic or earthquake event, I leverage the machine learning techniques to help recognizing seismic waveforms in microseismic or earthquake data. Because of the dependency of supervised machine learning algorithm on large volume of well-designed training data, I utilize an unsupervised machine learning algorithm to help cluster the time samples into two groups, that is, waveform points and non-waveform points. The fuzzy clustering algorithm has been demonstrated to be effective for such purpose. A group of synthetic, real microseismic and earthquake data sets with different levels of complexity show that the proposed method is much more robust than the state-of-the-art STA/LTA method in picking microseismic events, even in the case of moderately strong background noise.

  1. Data Clustering

    Science.gov (United States)

    Wagstaff, Kiri L.

    2012-03-01

    On obtaining a new data set, the researcher is immediately faced with the challenge of obtaining a high-level understanding from the observations. What does a typical item look like? What are the dominant trends? How many distinct groups are included in the data set, and how is each one characterized? Which observable values are common, and which rarely occur? Which items stand out as anomalies or outliers from the rest of the data? This challenge is exacerbated by the steady growth in data set size [11] as new instruments push into new frontiers of parameter space, via improvements in temporal, spatial, and spectral resolution, or by the desire to "fuse" observations from different modalities and instruments into a larger-picture understanding of the same underlying phenomenon. Data clustering algorithms provide a variety of solutions for this task. They can generate summaries, locate outliers, compress data, identify dense or sparse regions of feature space, and build data models. It is useful to note up front that "clusters" in this context refer to groups of items within some descriptive feature space, not (necessarily) to "galaxy clusters" which are dense regions in physical space. The goal of this chapter is to survey a variety of data clustering methods, with an eye toward their applicability to astronomical data analysis. In addition to improving the individual researcher’s understanding of a given data set, clustering has led directly to scientific advances, such as the discovery of new subclasses of stars [14] and gamma-ray bursts (GRBs) [38]. All clustering algorithms seek to identify groups within a data set that reflect some observed, quantifiable structure. Clustering is traditionally an unsupervised approach to data analysis, in the sense that it operates without any direct guidance about which items should be assigned to which clusters. There has been a recent trend in the clustering literature toward supporting semisupervised or constrained

  2. Hierarchical clustering of breast cancer methylomes revealed differentially methylated and expressed breast cancer genes.

    Directory of Open Access Journals (Sweden)

    I-Hsuan Lin

    Full Text Available Oncogenic transformation of normal cells often involves epigenetic alterations, including histone modification and DNA methylation. We conducted whole-genome bisulfite sequencing to determine the DNA methylomes of normal breast, fibroadenoma, invasive ductal carcinomas and MCF7. The emergence, disappearance, expansion and contraction of kilobase-sized hypomethylated regions (HMRs and the hypomethylation of the megabase-sized partially methylated domains (PMDs are the major forms of methylation changes observed in breast tumor samples. Hierarchical clustering of HMR revealed tumor-specific hypermethylated clusters and differential methylated enhancers specific to normal or breast cancer cell lines. Joint analysis of gene expression and DNA methylation data of normal breast and breast cancer cells identified differentially methylated and expressed genes associated with breast and/or ovarian cancers in cancer-specific HMR clusters. Furthermore, aberrant patterns of X-chromosome inactivation (XCI was found in breast cancer cell lines as well as breast tumor samples in the TCGA BRCA (breast invasive carcinoma dataset. They were characterized with differentially hypermethylated XIST promoter, reduced expression of XIST, and over-expression of hypomethylated X-linked genes. High expressions of these genes were significantly associated with lower survival rates in breast cancer patients. Comprehensive analysis of the normal and breast tumor methylomes suggests selective targeting of DNA methylation changes during breast cancer progression. The weak causal relationship between DNA methylation and gene expression observed in this study is evident of more complex role of DNA methylation in the regulation of gene expression in human epigenetics that deserves further investigation.

  3. An improved Pearson's correlation proximity-based hierarchical clustering for mining biological association between genes.

    Science.gov (United States)

    Booma, P M; Prabhakaran, S; Dhanalakshmi, R

    2014-01-01

    Microarray gene expression datasets has concerned great awareness among molecular biologist, statisticians, and computer scientists. Data mining that extracts the hidden and usual information from datasets fails to identify the most significant biological associations between genes. A search made with heuristic for standard biological process measures only the gene expression level, threshold, and response time. Heuristic search identifies and mines the best biological solution, but the association process was not efficiently addressed. To monitor higher rate of expression levels between genes, a hierarchical clustering model was proposed, where the biological association between genes is measured simultaneously using proximity measure of improved Pearson's correlation (PCPHC). Additionally, the Seed Augment algorithm adopts average linkage methods on rows and columns in order to expand a seed PCPHC model into a maximal global PCPHC (GL-PCPHC) model and to identify association between the clusters. Moreover, a GL-PCPHC applies pattern growing method to mine the PCPHC patterns. Compared to existing gene expression analysis, the PCPHC model achieves better performance. Experimental evaluations are conducted for GL-PCPHC model with standard benchmark gene expression datasets extracted from UCI repository and GenBank database in terms of execution time, size of pattern, significance level, biological association efficiency, and pattern quality.

  4. Correlates of Unsupervised Bathing of Infants: A Cross-Sectional Study

    Directory of Open Access Journals (Sweden)

    Tinneke M. J. Beirens

    2013-03-01

    Full Text Available Drowning represents the third leading cause of fatal unintentional injury in infants (0–1 years. The aim of this study is to investigate correlates of unsupervised bathing. This cross-sectional study included 1,410 parents with an infant. Parents completed a questionnaire regarding supervision during bathing, socio-demographic factors, and Protection Motivation Theory-constructs. To determine correlates of parents who leave their infant unsupervised, logistic regression analyses were performed. Of the parents, 6.2% left their child unsupervised in the bathtub. Parents with older children (OR 1.24; 95%CI 1.00–1.54 were more likely to leave their child unsupervised in the bathtub. First-time parents (OR 0.59; 95%CI 0.36–0.97 and non-Western migrant fathers (OR 0.18; 95%CI 0.05–0.63 were less likely to leave their child unsupervised in the bathtub. Furthermore, parents who perceived higher self-efficacy (OR 0.57; 95%CI 0.47–0.69, higher response efficacy (OR 0.34; 95%CI 0.24–0.48, and higher severity (OR 0.74; 95%CI 0.58–0.93 were less likely to leave their child unsupervised. Since young children are at great risk of drowning if supervision is absent, effective strategies for drowning prevention should be developed and evaluated. In the meantime, health care professionals should inform parents with regard to the importance of supervision during bathing.

  5. An evaluation of unsupervised and supervised learning algorithms for clustering landscape types in the United States

    Science.gov (United States)

    Wendel, Jochen; Buttenfield, Barbara P.; Stanislawski, Larry V.

    2016-01-01

    Knowledge of landscape type can inform cartographic generalization of hydrographic features, because landscape characteristics provide an important geographic context that affects variation in channel geometry, flow pattern, and network configuration. Landscape types are characterized by expansive spatial gradients, lacking abrupt changes between adjacent classes; and as having a limited number of outliers that might confound classification. The US Geological Survey (USGS) is exploring methods to automate generalization of features in the National Hydrography Data set (NHD), to associate specific sequences of processing operations and parameters with specific landscape characteristics, thus obviating manual selection of a unique processing strategy for every NHD watershed unit. A chronology of methods to delineate physiographic regions for the United States is described, including a recent maximum likelihood classification based on seven input variables. This research compares unsupervised and supervised algorithms applied to these seven input variables, to evaluate and possibly refine the recent classification. Evaluation metrics for unsupervised methods include the Davies–Bouldin index, the Silhouette index, and the Dunn index as well as quantization and topographic error metrics. Cross validation and misclassification rate analysis are used to evaluate supervised classification methods. The paper reports the comparative analysis and its impact on the selection of landscape regions. The compared solutions show problems in areas of high landscape diversity. There is some indication that additional input variables, additional classes, or more sophisticated methods can refine the existing classification.

  6. Unsupervised Learning and Pattern Recognition of Biological Data Structures with Density Functional Theory and Machine Learning.

    Science.gov (United States)

    Chen, Chien-Chang; Juan, Hung-Hui; Tsai, Meng-Yuan; Lu, Henry Horng-Shing

    2018-01-11

    By introducing the methods of machine learning into the density functional theory, we made a detour for the construction of the most probable density function, which can be estimated by learning relevant features from the system of interest. Using the properties of universal functional, the vital core of density functional theory, the most probable cluster numbers and the corresponding cluster boundaries in a studying system can be simultaneously and automatically determined and the plausibility is erected on the Hohenberg-Kohn theorems. For the method validation and pragmatic applications, interdisciplinary problems from physical to biological systems were enumerated. The amalgamation of uncharged atomic clusters validated the unsupervised searching process of the cluster numbers and the corresponding cluster boundaries were exhibited likewise. High accurate clustering results of the Fisher's iris dataset showed the feasibility and the flexibility of the proposed scheme. Brain tumor detections from low-dimensional magnetic resonance imaging datasets and segmentations of high-dimensional neural network imageries in the Brainbow system were also used to inspect the method practicality. The experimental results exhibit the successful connection between the physical theory and the machine learning methods and will benefit the clinical diagnoses.

  7. A Novel Cluster Head Selection Algorithm Based on Fuzzy Clustering and Particle Swarm Optimization.

    Science.gov (United States)

    Ni, Qingjian; Pan, Qianqian; Du, Huimin; Cao, Cen; Zhai, Yuqing

    2017-01-01

    An important objective of wireless sensor network is to prolong the network life cycle, and topology control is of great significance for extending the network life cycle. Based on previous work, for cluster head selection in hierarchical topology control, we propose a solution based on fuzzy clustering preprocessing and particle swarm optimization. More specifically, first, fuzzy clustering algorithm is used to initial clustering for sensor nodes according to geographical locations, where a sensor node belongs to a cluster with a determined probability, and the number of initial clusters is analyzed and discussed. Furthermore, the fitness function is designed considering both the energy consumption and distance factors of wireless sensor network. Finally, the cluster head nodes in hierarchical topology are determined based on the improved particle swarm optimization. Experimental results show that, compared with traditional methods, the proposed method achieved the purpose of reducing the mortality rate of nodes and extending the network life cycle.

  8. Unsupervised laparoscopic appendicectomy by surgical trainees is safe and time-effective.

    Science.gov (United States)

    Wong, Kenneth; Duncan, Tristram; Pearson, Andrew

    2007-07-01

    Open appendicectomy is the traditional standard treatment for appendicitis. Laparoscopic appendicectomy is perceived as a procedure with greater potential for complications and longer operative times. This paper examines the hypothesis that unsupervised laparoscopic appendicectomy by surgical trainees is a safe and time-effective valid alternative. Medical records, operating theatre records and histopathology reports of all patients undergoing laparoscopic and open appendicectomy over a 15-month period in two hospitals within an area health service were retrospectively reviewed. Data were analysed to compare patient features, pathology findings, operative times, complications, readmissions and mortality between laparoscopic and open groups and between unsupervised surgical trainee operators versus consultant surgeon operators. A total of 143 laparoscopic and 222 open appendicectomies were reviewed. Unsupervised trainees performed 64% of the laparoscopic appendicectomies and 55% of the open appendicectomies. There were no significant differences in complication rates, readmissions, mortality and length of stay between laparoscopic and open appendicectomy groups or between trainee and consultant surgeon operators. Conversion rates (laparoscopic to open approach) were similar for trainees and consultants. Unsupervised senior surgical trainees did not take significantly longer to perform laparoscopic appendicectomy when compared to unsupervised trainee-performed open appendicectomy. Unsupervised laparoscopic appendicectomy by surgical trainees is safe and time-effective.

  9. Machine learning in APOGEE. Unsupervised spectral classification with K-means

    Science.gov (United States)

    Garcia-Dias, Rafael; Allende Prieto, Carlos; Sánchez Almeida, Jorge; Ordovás-Pascual, Ignacio

    2018-05-01

    Context. The volume of data generated by astronomical surveys is growing rapidly. Traditional analysis techniques in spectroscopy either demand intensive human interaction or are computationally expensive. In this scenario, machine learning, and unsupervised clustering algorithms in particular, offer interesting alternatives. The Apache Point Observatory Galactic Evolution Experiment (APOGEE) offers a vast data set of near-infrared stellar spectra, which is perfect for testing such alternatives. Aims: Our research applies an unsupervised classification scheme based on K-means to the massive APOGEE data set. We explore whether the data are amenable to classification into discrete classes. Methods: We apply the K-means algorithm to 153 847 high resolution spectra (R ≈ 22 500). We discuss the main virtues and weaknesses of the algorithm, as well as our choice of parameters. Results: We show that a classification based on normalised spectra captures the variations in stellar atmospheric parameters, chemical abundances, and rotational velocity, among other factors. The algorithm is able to separate the bulge and halo populations, and distinguish dwarfs, sub-giants, RC, and RGB stars. However, a discrete classification in flux space does not result in a neat organisation in the parameters' space. Furthermore, the lack of obvious groups in flux space causes the results to be fairly sensitive to the initialisation, and disrupts the efficiency of commonly-used methods to select the optimal number of clusters. Our classification is publicly available, including extensive online material associated with the APOGEE Data Release 12 (DR12). Conclusions: Our description of the APOGEE database can help greatly with the identification of specific types of targets for various applications. We find a lack of obvious groups in flux space, and identify limitations of the K-means algorithm in dealing with this kind of data. Full Tables B.1-B.4 are only available at the CDS via

  10. Enhancement of ELM by Clustering Discrimination Manifold Regularization and Multiobjective FOA for Semisupervised Classification

    OpenAIRE

    Qing Ye; Hao Pan; Changhua Liu

    2015-01-01

    A novel semisupervised extreme learning machine (ELM) with clustering discrimination manifold regularization (CDMR) framework named CDMR-ELM is proposed for semisupervised classification. By using unsupervised fuzzy clustering method, CDMR framework integrates clustering discrimination of both labeled and unlabeled data with twinning constraints regularization. Aiming at further improving the classification accuracy and efficiency, a new multiobjective fruit fly optimization algorithm (MOFOA)...

  11. A new application of hierarchical cluster analysis to investigate organic peaks in bulk mass spectra obtained with an Aerodyne Aerosol Mass Spectrometer

    Science.gov (United States)

    Middlebrook, A. M.; Marcolli, C.; Canagaratna, M. R.; Worsnop, D. R.; Bahreini, R.; de Gouw, J. A.; Warneke, C.; Goldan, P. D.; Kuster, W. C.; Williams, E. J.; Lerner, B. M.; Roberts, J. M.; Meagher, J. F.; Fehsenfeld, F. C.; Marchewka, M. L.; Bertman, S. B.

    2006-12-01

    We applied hierarchical cluster analysis to an Aerodyne aerosol mass spectrometer (AMS) bulk mass spectral dataset collected aboard the NOAA research vessel Ronald H. Brown during the 2002 New England Air Quality Study off the east coast of the United States. Emphasizing the organic peaks, the cluster analysis yielded a series of categories that are distinguishable with respect to their mass spectra and their occurrence as a function of time. The differences between the categories mainly arise from relative intensity changes rather than from the presence or absence of specific peaks. The most frequent category exhibits a strong signal at m/z 44 and represents oxidized organic matter probably originating from both anthropogenic as well as biogenic sources. On the basis of spectral and trace gas correlations, the second most common category with strong signals at m/z 29, 43, and 44 contains contributions from isoprene oxidation products. The third through the fifth most common categories have peak patterns characteristic of monoterpene oxidation products and were most frequently observed when air masses from monoterpene rich regions were sampled. Taken together, the second through the fifth most common categories represent on average 17% of the total organic mass that stems likely from biogenic sources during the ship's cruise. These numbers have to be viewed as lower limits since the most common category was attributed to anthropogenic sources for this calculation. The cluster analysis was also very effective in identifying a few contaminated mass spectra that were not removed during pre-processing. This study demonstrates that hierarchical clustering is a useful tool to analyze the complex patterns of the organic peaks in bulk aerosol mass spectra from a field study.

  12. Unsupervised text mining methods for literature analysis: a case study for Thomas Pynchon's V.

    Directory of Open Access Journals (Sweden)

    Christos Iraklis Tsatsoulis

    2013-08-01

    Full Text Available We investigate the use of unsupervised text mining methods for the analysis of prose literature works, using Thomas Pynchon's novel 'V'. as a case study. Our results suggest that such methods may be employed to reveal meaningful information regarding the novel’s structure. We report results using a wide variety of clustering algorithms, several distinct distance functions, and different visualization techniques. The application of a simple topic model is also demonstrated. We discuss the meaningfulness of our results along with the limitations of our approach, and we suggest some possible paths for further study.

  13. An unsupervised strategy for biomedical image segmentation

    Directory of Open Access Journals (Sweden)

    Roberto Rodríguez

    2010-09-01

    Full Text Available Roberto Rodríguez1, Rubén Hernández21Digital Signal Processing Group, Institute of Cybernetics, Mathematics, and Physics, Havana, Cuba; 2Interdisciplinary Professional Unit of Engineering and Advanced Technology, IPN, MexicoAbstract: Many segmentation techniques have been published, and some of them have been widely used in different application problems. Most of these segmentation techniques have been motivated by specific application purposes. Unsupervised methods, which do not assume any prior scene knowledge can be learned to help the segmentation process, and are obviously more challenging than the supervised ones. In this paper, we present an unsupervised strategy for biomedical image segmentation using an algorithm based on recursively applying mean shift filtering, where entropy is used as a stopping criterion. This strategy is proven with many real images, and a comparison is carried out with manual segmentation. With the proposed strategy, errors less than 20% for false positives and 0% for false negatives are obtained.Keywords: segmentation, mean shift, unsupervised segmentation, entropy

  14. Unsupervised image matching based on manifold alignment.

    Science.gov (United States)

    Pei, Yuru; Huang, Fengchun; Shi, Fuhao; Zha, Hongbin

    2012-08-01

    This paper challenges the issue of automatic matching between two image sets with similar intrinsic structures and different appearances, especially when there is no prior correspondence. An unsupervised manifold alignment framework is proposed to establish correspondence between data sets by a mapping function in the mutual embedding space. We introduce a local similarity metric based on parameterized distance curves to represent the connection of one point with the rest of the manifold. A small set of valid feature pairs can be found without manual interactions by matching the distance curve of one manifold with the curve cluster of the other manifold. To avoid potential confusions in image matching, we propose an extended affine transformation to solve the nonrigid alignment in the embedding space. The comparatively tight alignments and the structure preservation can be obtained simultaneously. The point pairs with the minimum distance after alignment are viewed as the matchings. We apply manifold alignment to image set matching problems. The correspondence between image sets of different poses, illuminations, and identities can be established effectively by our approach.

  15. AUTOMATED UNSUPERVISED CLASSIFICATION OF THE SLOAN DIGITAL SKY SURVEY STELLAR SPECTRA USING k-MEANS CLUSTERING

    Energy Technology Data Exchange (ETDEWEB)

    Sanchez Almeida, J.; Allende Prieto, C., E-mail: jos@iac.es, E-mail: callende@iac.es [Instituto de Astrofisica de Canarias, E-38205 La Laguna, Tenerife (Spain)

    2013-01-20

    Large spectroscopic surveys require automated methods of analysis. This paper explores the use of k-means clustering as a tool for automated unsupervised classification of massive stellar spectral catalogs. The classification criteria are defined by the data and the algorithm, with no prior physical framework. We work with a representative set of stellar spectra associated with the Sloan Digital Sky Survey (SDSS) SEGUE and SEGUE-2 programs, which consists of 173,390 spectra from 3800 to 9200 A sampled on 3849 wavelengths. We classify the original spectra as well as the spectra with the continuum removed. The second set only contains spectral lines, and it is less dependent on uncertainties of the flux calibration. The classification of the spectra with continuum renders 16 major classes. Roughly speaking, stars are split according to their colors, with enough finesse to distinguish dwarfs from giants of the same effective temperature, but with difficulties to separate stars with different metallicities. There are classes corresponding to particular MK types, intrinsically blue stars, dust-reddened, stellar systems, and also classes collecting faulty spectra. Overall, there is no one-to-one correspondence between the classes we derive and the MK types. The classification of spectra without continuum renders 13 classes, the color separation is not so sharp, but it distinguishes stars of the same effective temperature and different metallicities. Some classes thus obtained present a fairly small range of physical parameters (200 K in effective temperature, 0.25 dex in surface gravity, and 0.35 dex in metallicity), so that the classification can be used to estimate the main physical parameters of some stars at a minimum computational cost. We also analyze the outliers of the classification. Most of them turn out to be failures of the reduction pipeline, but there are also high redshift QSOs, multiple stellar systems, dust-reddened stars, galaxies, and, finally, odd

  16. Concept formation knowledge and experience in unsupervised learning

    CERN Document Server

    Fisher, Douglas H; Langley, Pat

    1991-01-01

    Concept Formation: Knowledge and Experience in Unsupervised Learning presents the interdisciplinary interaction between machine learning and cognitive psychology on unsupervised incremental methods. This book focuses on measures of similarity, strategies for robust incremental learning, and the psychological consistency of various approaches.Organized into three parts encompassing 15 chapters, this book begins with an overview of inductive concept learning in machine learning and psychology, with emphasis on issues that distinguish concept formation from more prevalent supervised methods and f

  17. Unsupervised text mining for assessing and augmenting GWAS results.

    Science.gov (United States)

    Ailem, Melissa; Role, François; Nadif, Mohamed; Demenais, Florence

    2016-04-01

    Text mining can assist in the analysis and interpretation of large-scale biomedical data, helping biologists to quickly and cheaply gain confirmation of hypothesized relationships between biological entities. We set this question in the context of genome-wide association studies (GWAS), an actively emerging field that contributed to identify many genes associated with multifactorial diseases. These studies allow to identify groups of genes associated with the same phenotype, but provide no information about the relationships between these genes. Therefore, our objective is to leverage unsupervised text mining techniques using text-based cosine similarity comparisons and clustering applied to candidate and random gene vectors, in order to augment the GWAS results. We propose a generic framework which we used to characterize the relationships between 10 genes reported associated with asthma by a previous GWAS. The results of this experiment showed that the similarities between these 10 genes were significantly stronger than would be expected by chance (one-sided p-value<0.01). The clustering of observed and randomly selected gene also allowed to generate hypotheses about potential functional relationships between these genes and thus contributed to the discovery of new candidate genes for asthma. Copyright © 2016 Elsevier Inc. All rights reserved.

  18. Brightest Cluster Galaxies in REXCESS Clusters

    Science.gov (United States)

    Haarsma, Deborah B.; Leisman, L.; Bruch, S.; Donahue, M.

    2009-01-01

    Most galaxy clusters contain a Brightest Cluster Galaxy (BCG) which is larger than the other cluster ellipticals and has a more extended profile. In the hierarchical model, the BCG forms through many galaxy mergers in the crowded center of the cluster, and thus its properties give insight into the assembly of the cluster as a whole. In this project, we are working with the Representative XMM-Newton Cluster Structure Survey (REXCESS) team (Boehringer et al 2007) to study BCGs in 33 X-ray luminous galaxy clusters, 0.055 < z < 0.183. We are imaging the BCGs in R band at the Southern Observatory for Astrophysical Research (SOAR) in Chile. In this poster, we discuss our methods and give preliminary measurements of the BCG magnitudes, morphology, and stellar mass. We compare these BCG properties with the properties of their host clusters, particularly of the X-ray emitting gas.

  19. Cluster evolution

    International Nuclear Information System (INIS)

    Schaeffer, R.

    1987-01-01

    The galaxy and cluster luminosity functions are constructed from a model of the mass distribution based on hierarchical clustering at an epoch where the matter distribution is non-linear. These luminosity functions are seen to reproduce the present distribution of objects as can be inferred from the observations. They can be used to deduce the redshift dependence of the cluster distribution and to extrapolate the observations towards the past. The predicted evolution of the cluster distribution is quite strong, although somewhat less rapid than predicted by the linear theory

  20. CytoCluster: A Cytoscape Plugin for Cluster Analysis and Visualization of Biological Networks.

    Science.gov (United States)

    Li, Min; Li, Dongyan; Tang, Yu; Wu, Fangxiang; Wang, Jianxin

    2017-08-31

    Nowadays, cluster analysis of biological networks has become one of the most important approaches to identifying functional modules as well as predicting protein complexes and network biomarkers. Furthermore, the visualization of clustering results is crucial to display the structure of biological networks. Here we present CytoCluster, a cytoscape plugin integrating six clustering algorithms, HC-PIN (Hierarchical Clustering algorithm in Protein Interaction Networks), OH-PIN (identifying Overlapping and Hierarchical modules in Protein Interaction Networks), IPCA (Identifying Protein Complex Algorithm), ClusterONE (Clustering with Overlapping Neighborhood Expansion), DCU (Detecting Complexes based on Uncertain graph model), IPC-MCE (Identifying Protein Complexes based on Maximal Complex Extension), and BinGO (the Biological networks Gene Ontology) function. Users can select different clustering algorithms according to their requirements. The main function of these six clustering algorithms is to detect protein complexes or functional modules. In addition, BinGO is used to determine which Gene Ontology (GO) categories are statistically overrepresented in a set of genes or a subgraph of a biological network. CytoCluster can be easily expanded, so that more clustering algorithms and functions can be added to this plugin. Since it was created in July 2013, CytoCluster has been downloaded more than 9700 times in the Cytoscape App store and has already been applied to the analysis of different biological networks. CytoCluster is available from http://apps.cytoscape.org/apps/cytocluster.

  1. Relation between financial market structure and the real economy: comparison between clustering methods.

    Science.gov (United States)

    Musmeci, Nicoló; Aste, Tomaso; Di Matteo, T

    2015-01-01

    We quantify the amount of information filtered by different hierarchical clustering methods on correlations between stock returns comparing the clustering structure with the underlying industrial activity classification. We apply, for the first time to financial data, a novel hierarchical clustering approach, the Directed Bubble Hierarchical Tree and we compare it with other methods including the Linkage and k-medoids. By taking the industrial sector classification of stocks as a benchmark partition, we evaluate how the different methods retrieve this classification. The results show that the Directed Bubble Hierarchical Tree can outperform other methods, being able to retrieve more information with fewer clusters. Moreover,we show that the economic information is hidden at different levels of the hierarchical structures depending on the clustering method. The dynamical analysis on a rolling window also reveals that the different methods show different degrees of sensitivity to events affecting financial markets, like crises. These results can be of interest for all the applications of clustering methods to portfolio optimization and risk hedging [corrected].

  2. Relation between financial market structure and the real economy: comparison between clustering methods.

    Directory of Open Access Journals (Sweden)

    Nicoló Musmeci

    Full Text Available We quantify the amount of information filtered by different hierarchical clustering methods on correlations between stock returns comparing the clustering structure with the underlying industrial activity classification. We apply, for the first time to financial data, a novel hierarchical clustering approach, the Directed Bubble Hierarchical Tree and we compare it with other methods including the Linkage and k-medoids. By taking the industrial sector classification of stocks as a benchmark partition, we evaluate how the different methods retrieve this classification. The results show that the Directed Bubble Hierarchical Tree can outperform other methods, being able to retrieve more information with fewer clusters. Moreover,we show that the economic information is hidden at different levels of the hierarchical structures depending on the clustering method. The dynamical analysis on a rolling window also reveals that the different methods show different degrees of sensitivity to events affecting financial markets, like crises. These results can be of interest for all the applications of clustering methods to portfolio optimization and risk hedging [corrected].

  3. Audio-based, unsupervised machine learning reveals cyclic changes in earthquake mechanisms in the Geysers geothermal field, California

    Science.gov (United States)

    Holtzman, B. K.; Paté, A.; Paisley, J.; Waldhauser, F.; Repetto, D.; Boschi, L.

    2017-12-01

    The earthquake process reflects complex interactions of stress, fracture and frictional properties. New machine learning methods reveal patterns in time-dependent spectral properties of seismic signals and enable identification of changes in faulting processes. Our methods are based closely on those developed for music information retrieval and voice recognition, using the spectrogram instead of the waveform directly. Unsupervised learning involves identification of patterns based on differences among signals without any additional information provided to the algorithm. Clustering of 46,000 earthquakes of $0.3

  4. Automated age-related macular degeneration classification in OCT using unsupervised feature learning

    Science.gov (United States)

    Venhuizen, Freerk G.; van Ginneken, Bram; Bloemen, Bart; van Grinsven, Mark J. J. P.; Philipsen, Rick; Hoyng, Carel; Theelen, Thomas; Sánchez, Clara I.

    2015-03-01

    Age-related Macular Degeneration (AMD) is a common eye disorder with high prevalence in elderly people. The disease mainly affects the central part of the retina, and could ultimately lead to permanent vision loss. Optical Coherence Tomography (OCT) is becoming the standard imaging modality in diagnosis of AMD and the assessment of its progression. However, the evaluation of the obtained volumetric scan is time consuming, expensive and the signs of early AMD are easy to miss. In this paper we propose a classification method to automatically distinguish AMD patients from healthy subjects with high accuracy. The method is based on an unsupervised feature learning approach, and processes the complete image without the need for an accurate pre-segmentation of the retina. The method can be divided in two steps: an unsupervised clustering stage that extracts a set of small descriptive image patches from the training data, and a supervised training stage that uses these patches to create a patch occurrence histogram for every image on which a random forest classifier is trained. Experiments using 384 volume scans show that the proposed method is capable of identifying AMD patients with high accuracy, obtaining an area under the Receiver Operating Curve of 0:984. Our method allows for a quick and reliable assessment of the presence of AMD pathology in OCT volume scans without the need for accurate layer segmentation algorithms.

  5. Automated segmentation of white matter fiber bundles using diffusion tensor imaging data and a new density based clustering algorithm.

    Science.gov (United States)

    Kamali, Tahereh; Stashuk, Daniel

    2016-10-01

    Robust and accurate segmentation of brain white matter (WM) fiber bundles assists in diagnosing and assessing progression or remission of neuropsychiatric diseases such as schizophrenia, autism and depression. Supervised segmentation methods are infeasible in most applications since generating gold standards is too costly. Hence, there is a growing interest in designing unsupervised methods. However, most conventional unsupervised methods require the number of clusters be known in advance which is not possible in most applications. The purpose of this study is to design an unsupervised segmentation algorithm for brain white matter fiber bundles which can automatically segment fiber bundles using intrinsic diffusion tensor imaging data information without considering any prior information or assumption about data distributions. Here, a new density based clustering algorithm called neighborhood distance entropy consistency (NDEC), is proposed which discovers natural clusters within data by simultaneously utilizing both local and global density information. The performance of NDEC is compared with other state of the art clustering algorithms including chameleon, spectral clustering, DBSCAN and k-means using Johns Hopkins University publicly available diffusion tensor imaging data. The performance of NDEC and other employed clustering algorithms were evaluated using dice ratio as an external evaluation criteria and density based clustering validation (DBCV) index as an internal evaluation metric. Across all employed clustering algorithms, NDEC obtained the highest average dice ratio (0.94) and DBCV value (0.71). NDEC can find clusters with arbitrary shapes and densities and consequently can be used for WM fiber bundle segmentation where there is no distinct boundary between various bundles. NDEC may also be used as an effective tool in other pattern recognition and medical diagnostic systems in which discovering natural clusters within data is a necessity. Copyright

  6. Unsupervised Categorization in a Sample of Children with Autism Spectrum Disorders

    Science.gov (United States)

    Edwards, Darren J.; Perlman, Amotz; Reed, Phil

    2012-01-01

    Studies of supervised Categorization have demonstrated limited Categorization performance in participants with autism spectrum disorders (ASD), however little research has been conducted regarding unsupervised Categorization in this population. This study explored unsupervised Categorization using two stimulus sets that differed in their…

  7. Unsupervised Learning of Structural Representation of Percussive Audio Using a Hierarchical Dirichlet Process Hidden Markov Model

    DEFF Research Database (Denmark)

    Antich, Jose Luis Diez; Paterna, Mattia; Marxer, Richard

    2016-01-01

    to jointly estimate the optimal number of sound clusters, to cluster the blocks, and to estimate the transition probabilities between clusters. The result is a segmentation of the input into a sequence of symbols (typically corresponding to hits of hi-hat, snare, bass, cymbal, etc.) that can be evaluated...

  8. Application of hierarchical clustering method to classify of space-time rainfall patterns

    Science.gov (United States)

    Yu, Hwa-Lung; Chang, Tu-Je

    2010-05-01

    Understanding the local precipitation patterns is essential to the water resources management and flooding mitigation. The precipitation patterns can vary in space and time depending upon the factors from different spatial scales such as local topological changes and macroscopic atmospheric circulation. The spatiotemporal variation of precipitation in Taiwan is significant due to its complex terrain and its location at west pacific and subtropical area, where is the boundary between the pacific ocean and Asia continent with the complex interactions among the climatic processes. This study characterizes local-scale precipitation patterns by classifying the historical space-time precipitation records. We applied the hierarchical ascending clustering method to analyze the precipitation records from 1960 to 2008 at the six rainfall stations located in Lan-yang catchment at the northeast of the island. Our results identify the four primary space-time precipitation types which may result from distinct driving forces from the changes of atmospheric variables and topology at different space-time scales. This study also presents an important application of the statistical downscaling to combine large-scale upper-air circulation with local space-time precipitation patterns.

  9. CBHRP: A Cluster Based Routing Protocol for Wireless Sensor Network

    OpenAIRE

    Rashed, M. G.; Kabir, M. Hasnat; Rahim, M. Sajjadur; Ullah, Sk. Enayet

    2012-01-01

    A new two layer hierarchical routing protocol called Cluster Based Hierarchical Routing Protocol (CBHRP) is proposed in this paper. It is an extension of LEACH routing protocol. We introduce cluster head-set idea for cluster-based routing where several clusters are formed with the deployed sensors to collect information from target field. On rotation basis, a head-set member receives data from the neighbor nodes and transmits the aggregated results to the distance base station. This protocol ...

  10. NOVEL CONTEXT-AWARE CLUSTERING WITH HIERARCHICAL ADDRESSING (CCHA) FOR THE INTERNET OF THINGS (IoT)

    DEFF Research Database (Denmark)

    Mahalle, Parikshit N.; Prasad, Neeli R.; Prasad, Ramjee

    2013-01-01

    As computing technology becomes more tightly coupled into dynamic and mobile world of the Internet of Things (IoT), security mechanism becomes more stringent, less flexible and intrusive. Scalability issue in the IoT makes Identity Management (IdM) of ubiquitous things more challenging. Forming ad......-hoc network, interaction between these nomadic devices to provide seamless service extend the need of new identi-ties to the things, addressing and IdM in the IoT. New identities and identifier format to alleviate the perfor-mance issue is introduced in this paper. This paper pre-sents novel Context......-aware Clustering with Hierarchical Addressing (CCHA) scheme for the things with new identifier format. Simulation results shows that CCHA achieves better performance with less energy expendi-ture, less end-to-end delay and more throughput. Results also show that CCHA significantly reduces the failure probability...

  11. Using Hierarchical Time Series Clustering Algorithm and Wavelet Classifier for Biometric Voice Classification

    Directory of Open Access Journals (Sweden)

    Simon Fong

    2012-01-01

    Full Text Available Voice biometrics has a long history in biosecurity applications such as verification and identification based on characteristics of the human voice. The other application called voice classification which has its important role in grouping unlabelled voice samples, however, has not been widely studied in research. Lately voice classification is found useful in phone monitoring, classifying speakers’ gender, ethnicity and emotion states, and so forth. In this paper, a collection of computational algorithms are proposed to support voice classification; the algorithms are a combination of hierarchical clustering, dynamic time wrap transform, discrete wavelet transform, and decision tree. The proposed algorithms are relatively more transparent and interpretable than the existing ones, though many techniques such as Artificial Neural Networks, Support Vector Machine, and Hidden Markov Model (which inherently function like a black box have been applied for voice verification and voice identification. Two datasets, one that is generated synthetically and the other one empirically collected from past voice recognition experiment, are used to verify and demonstrate the effectiveness of our proposed voice classification algorithm.

  12. Modular networks with hierarchical organization

    Indian Academy of Sciences (India)

    Several networks occurring in real life have modular structures that are arranged in a hierarchical fashion. In this paper, we have proposed a model for such networks, using a stochastic generation method. Using this model we show that, the scaling relation between the clustering and degree of the nodes is not a necessary ...

  13. Functional annotation of hierarchical modularity.

    Directory of Open Access Journals (Sweden)

    Kanchana Padmanabhan

    Full Text Available In biological networks of molecular interactions in a cell, network motifs that are biologically relevant are also functionally coherent, or form functional modules. These functionally coherent modules combine in a hierarchical manner into larger, less cohesive subsystems, thus revealing one of the essential design principles of system-level cellular organization and function-hierarchical modularity. Arguably, hierarchical modularity has not been explicitly taken into consideration by most, if not all, functional annotation systems. As a result, the existing methods would often fail to assign a statistically significant functional coherence score to biologically relevant molecular machines. We developed a methodology for hierarchical functional annotation. Given the hierarchical taxonomy of functional concepts (e.g., Gene Ontology and the association of individual genes or proteins with these concepts (e.g., GO terms, our method will assign a Hierarchical Modularity Score (HMS to each node in the hierarchy of functional modules; the HMS score and its p-value measure functional coherence of each module in the hierarchy. While existing methods annotate each module with a set of "enriched" functional terms in a bag of genes, our complementary method provides the hierarchical functional annotation of the modules and their hierarchically organized components. A hierarchical organization of functional modules often comes as a bi-product of cluster analysis of gene expression data or protein interaction data. Otherwise, our method will automatically build such a hierarchy by directly incorporating the functional taxonomy information into the hierarchy search process and by allowing multi-functional genes to be part of more than one component in the hierarchy. In addition, its underlying HMS scoring metric ensures that functional specificity of the terms across different levels of the hierarchical taxonomy is properly treated. We have evaluated our

  14. BULGELESS GIANT GALAXIES CHALLENGE OUR PICTURE OF GALAXY FORMATION BY HIERARCHICAL CLUSTERING ,

    International Nuclear Information System (INIS)

    Kormendy, John; Cornell, Mark E.; Drory, Niv; Bender, Ralf

    2010-01-01

    To better understand the prevalence of bulgeless galaxies in the nearby field, we dissect giant Sc-Scd galaxies with Hubble Space Telescope (HST) photometry and Hobby-Eberly Telescope (HET) spectroscopy. We use the HET High Resolution Spectrograph (resolution R ≡ λ/FWHM ≅ 15, 000) to measure stellar velocity dispersions in the nuclear star clusters and (pseudo)bulges of the pure-disk galaxies M 33, M 101, NGC 3338, NGC 3810, NGC 6503, and NGC 6946. The dispersions range from 20 ± 1 km s -1 in the nucleus of M 33 to 78 ± 2 km s -1 in the pseudobulge of NGC 3338. We use HST archive images to measure the brightness profiles of the nuclei and (pseudo)bulges in M 101, NGC 6503, and NGC 6946 and hence to estimate their masses. The results imply small mass-to-light ratios consistent with young stellar populations. These observations lead to two conclusions. (1) Upper limits on the masses of any supermassive black holes are M . ∼ 6 M sun in M 101 and M . ∼ 6 M sun in NGC 6503. (2) We show that the above galaxies contain only tiny pseudobulges that make up ∼ circ > 150 km s -1 , including M 101, NGC 6946, IC 342, and our Galaxy, show no evidence for a classical bulge. Four may contain small classical bulges that contribute 5%-12% of the light of the galaxy. Only four of the 19 giant galaxies are ellipticals or have classical bulges that contribute ∼1/3 of the galaxy light. We conclude that pure-disk galaxies are far from rare. It is hard to understand how bulgeless galaxies could form as the quiescent tail of a distribution of merger histories. Recognition of pseudobulges makes the biggest problem with cold dark matter galaxy formation more acute: How can hierarchical clustering make so many giant, pure-disk galaxies with no evidence for merger-built bulges? Finally, we emphasize that this problem is a strong function of environment: the Virgo cluster is not a puzzle, because more than 2/3 of its stellar mass is in merger remnants.

  15. BioCluster: Tool for Identification and Clustering of Enterobacteriaceae Based on Biochemical Data

    Directory of Open Access Journals (Sweden)

    Ahmed Abdullah

    2015-06-01

    Full Text Available Presumptive identification of different Enterobacteriaceae species is routinely achieved based on biochemical properties. Traditional practice includes manual comparison of each biochemical property of the unknown sample with known reference samples and inference of its identity based on the maximum similarity pattern with the known samples. This process is labor-intensive, time-consuming, error-prone, and subjective. Therefore, automation of sorting and similarity in calculation would be advantageous. Here we present a MATLAB-based graphical user interface (GUI tool named BioCluster. This tool was designed for automated clustering and identification of Enterobacteriaceae based on biochemical test results. In this tool, we used two types of algorithms, i.e., traditional hierarchical clustering (HC and the Improved Hierarchical Clustering (IHC, a modified algorithm that was developed specifically for the clustering and identification of Enterobacteriaceae species. IHC takes into account the variability in result of 1–47 biochemical tests within this Enterobacteriaceae family. This tool also provides different options to optimize the clustering in a user-friendly way. Using computer-generated synthetic data and some real data, we have demonstrated that BioCluster has high accuracy in clustering and identifying enterobacterial species based on biochemical test data. This tool can be freely downloaded at http://microbialgen.du.ac.bd/biocluster/.

  16. Clustering Dycom

    KAUST Repository

    Minku, Leandro L.

    2017-10-06

    Background: Software Effort Estimation (SEE) can be formulated as an online learning problem, where new projects are completed over time and may become available for training. In this scenario, a Cross-Company (CC) SEE approach called Dycom can drastically reduce the number of Within-Company (WC) projects needed for training, saving the high cost of collecting such training projects. However, Dycom relies on splitting CC projects into different subsets in order to create its CC models. Such splitting can have a significant impact on Dycom\\'s predictive performance. Aims: This paper investigates whether clustering methods can be used to help finding good CC splits for Dycom. Method: Dycom is extended to use clustering methods for creating the CC subsets. Three different clustering methods are investigated, namely Hierarchical Clustering, K-Means, and Expectation-Maximisation. Clustering Dycom is compared against the original Dycom with CC subsets of different sizes, based on four SEE databases. A baseline WC model is also included in the analysis. Results: Clustering Dycom with K-Means can potentially help to split the CC projects, managing to achieve similar or better predictive performance than Dycom. However, K-Means still requires the number of CC subsets to be pre-defined, and a poor choice can negatively affect predictive performance. EM enables Dycom to automatically set the number of CC subsets while still maintaining or improving predictive performance with respect to the baseline WC model. Clustering Dycom with Hierarchical Clustering did not offer significant advantage in terms of predictive performance. Conclusion: Clustering methods can be an effective way to automatically generate Dycom\\'s CC subsets.

  17. Evaluation of Hierarchical Clustering Algorithms for Document Datasets

    National Research Council Canada - National Science Library

    Zhao, Ying; Karypis, George

    2002-01-01

    Fast and high-quality document clustering algorithms play an important role in providing intuitive navigation and browsing mechanisms by organizing large amounts of information into a small number of meaningful clusters...

  18. DAFi: A directed recursive data filtering and clustering approach for improving and interpreting data clustering identification of cell populations from polychromatic flow cytometry data.

    Science.gov (United States)

    Lee, Alexandra J; Chang, Ivan; Burel, Julie G; Lindestam Arlehamn, Cecilia S; Mandava, Aishwarya; Weiskopf, Daniela; Peters, Bjoern; Sette, Alessandro; Scheuermann, Richard H; Qian, Yu

    2018-04-17

    Computational methods for identification of cell populations from polychromatic flow cytometry data are changing the paradigm of cytometry bioinformatics. Data clustering is the most common computational approach to unsupervised identification of cell populations from multidimensional cytometry data. However, interpretation of the identified data clusters is labor-intensive. Certain types of user-defined cell populations are also difficult to identify by fully automated data clustering analysis. Both are roadblocks before a cytometry lab can adopt the data clustering approach for cell population identification in routine use. We found that combining recursive data filtering and clustering with constraints converted from the user manual gating strategy can effectively address these two issues. We named this new approach DAFi: Directed Automated Filtering and Identification of cell populations. Design of DAFi preserves the data-driven characteristics of unsupervised clustering for identifying novel cell subsets, but also makes the results interpretable to experimental scientists through mapping and merging the multidimensional data clusters into the user-defined two-dimensional gating hierarchy. The recursive data filtering process in DAFi helped identify small data clusters which are otherwise difficult to resolve by a single run of the data clustering method due to the statistical interference of the irrelevant major clusters. Our experiment results showed that the proportions of the cell populations identified by DAFi, while being consistent with those by expert centralized manual gating, have smaller technical variances across samples than those from individual manual gating analysis and the nonrecursive data clustering analysis. Compared with manual gating segregation, DAFi-identified cell populations avoided the abrupt cut-offs on the boundaries. DAFi has been implemented to be used with multiple data clustering methods including K-means, FLOCK, FlowSOM, and

  19. Cyclist–motorist crash patterns in Denmark: A latent class clustering approach

    DEFF Research Database (Denmark)

    Kaplan, Sigal; Prato, Carlo Giacomo

    2013-01-01

    to prioritize safety issues and to devise efficient preventive measures. Method: The current study focused on cyclist–motorist crashes that occurred in Denmark during the period between 2007 and 2011. To uncover crash patterns, the current analysis applied latent class clustering, an unsupervised probabilistic...

  20. bcl::Cluster : A method for clustering biological molecules coupled with visualization in the Pymol Molecular Graphics System.

    Science.gov (United States)

    Alexander, Nathan; Woetzel, Nils; Meiler, Jens

    2011-02-01

    Clustering algorithms are used as data analysis tools in a wide variety of applications in Biology. Clustering has become especially important in protein structure prediction and virtual high throughput screening methods. In protein structure prediction, clustering is used to structure the conformational space of thousands of protein models. In virtual high throughput screening, databases with millions of drug-like molecules are organized by structural similarity, e.g. common scaffolds. The tree-like dendrogram structure obtained from hierarchical clustering can provide a qualitative overview of the results, which is important for focusing detailed analysis. However, in practice it is difficult to relate specific components of the dendrogram directly back to the objects of which it is comprised and to display all desired information within the two dimensions of the dendrogram. The current work presents a hierarchical agglomerative clustering method termed bcl::Cluster. bcl::Cluster utilizes the Pymol Molecular Graphics System to graphically depict dendrograms in three dimensions. This allows simultaneous display of relevant biological molecules as well as additional information about the clusters and the members comprising them.

  1. Shadow detection and removal in RGB VHR images for land use unsupervised classification

    Science.gov (United States)

    Movia, A.; Beinat, A.; Crosilla, F.

    2016-09-01

    Nowadays, high resolution aerial images are widely available thanks to the diffusion of advanced technologies such as UAVs (Unmanned Aerial Vehicles) and new satellite missions. Although these developments offer new opportunities for accurate land use analysis and change detection, cloud and terrain shadows actually limit benefits and possibilities of modern sensors. Focusing on the problem of shadow detection and removal in VHR color images, the paper proposes new solutions and analyses how they can enhance common unsupervised classification procedures for identifying land use classes related to the CO2 absorption. To this aim, an improved fully automatic procedure has been developed for detecting image shadows using exclusively RGB color information, and avoiding user interaction. Results show a significant accuracy enhancement with respect to similar methods using RGB based indexes. Furthermore, novel solutions derived from Procrustes analysis have been applied to remove shadows and restore brightness in the images. In particular, two methods implementing the so called "anisotropic Procrustes" and the "not-centered oblique Procrustes" algorithms have been developed and compared with the linear correlation correction method based on the Cholesky decomposition. To assess how shadow removal can enhance unsupervised classifications, results obtained with classical methods such as k-means, maximum likelihood, and self-organizing maps, have been compared to each other and with a supervised clustering procedure.

  2. Competitive cluster growth in complex networks.

    Science.gov (United States)

    Moreira, André A; Paula, Demétrius R; Costa Filho, Raimundo N; Andrade, José S

    2006-06-01

    In this work we propose an idealized model for competitive cluster growth in complex networks. Each cluster can be thought of as a fraction of a community that shares some common opinion. Our results show that the cluster size distribution depends on the particular choice for the topology of the network of contacts among the agents. As an application, we show that the cluster size distributions obtained when the growth process is performed on hierarchical networks, e.g., the Apollonian network, have a scaling form similar to what has been observed for the distribution of a number of votes in an electoral process. We suggest that this similarity may be due to the fact that social networks involved in the electoral process may also possess an underlining hierarchical structure.

  3. Chemical Fingerprint and Quantitative Analysis for the Quality Evaluation of Platycladi cacumen by Ultra-performance Liquid Chromatography Coupled with Hierarchical Cluster Analysis.

    Science.gov (United States)

    Shan, Mingqiu; Li, Sam Fong Yau; Yu, Sheng; Qian, Yan; Guo, Shuchen; Zhang, Li; Ding, Anwei

    2018-01-01

    Platycladi cacumen (dried twigs and leaves of Platycladus orientalis (L.) Franco) is a frequently utilized Chinese medicinal herb. To evaluate the quality of the phytomedcine, an ultra-performance liquid chromatographic method with diode array detection was established for chemical fingerprinting and quantitative analysis. In this study, 27 batches of P. cacumen from different regions were collected for analysis. A chemical fingerprint with 20 common peaks was obtained using Similarity Evaluation System for Chromatographic Fingerprint of Traditional Chinese Medicine (Version 2004A). Among these 20 components, seven flavonoids (myricitrin, isoquercitrin, quercitrin, afzelin, cupressuflavone, amentoflavone and hinokiflavone) were identified and determined simultaneously. In the method validation, the seven analytes showed good regressions (R ≥ 0.9995) within linear ranges and good recoveries from 96.4% to 103.3%. Furthermore, with the contents of these seven flavonoids, hierarchical clustering analysis was applied to distinguish the 27 batches into five groups. The chemometric results showed that these groups were almost consistent with geographical positions and climatic conditions of the production regions. Integrating fingerprint analysis, simultaneous determination and hierarchical clustering analysis, the established method is rapid, sensitive, accurate and readily applicable, and also provides a significant foundation for quality control of P. cacumen efficiently. © The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

  4. Kernel method for clustering based on optimal target vector

    International Nuclear Information System (INIS)

    Angelini, Leonardo; Marinazzo, Daniele; Pellicoro, Mario; Stramaglia, Sebastiano

    2006-01-01

    We introduce Ising models, suitable for dichotomic clustering, with couplings that are (i) both ferro- and anti-ferromagnetic (ii) depending on the whole data-set and not only on pairs of samples. Couplings are determined exploiting the notion of optimal target vector, here introduced, a link between kernel supervised and unsupervised learning. The effectiveness of the method is shown in the case of the well-known iris data-set and in benchmarks of gene expression levels, where it works better than existing methods for dichotomic clustering

  5. Algorithms of maximum likelihood data clustering with applications

    Science.gov (United States)

    Giada, Lorenzo; Marsili, Matteo

    2002-12-01

    We address the problem of data clustering by introducing an unsupervised, parameter-free approach based on maximum likelihood principle. Starting from the observation that data sets belonging to the same cluster share a common information, we construct an expression for the likelihood of any possible cluster structure. The likelihood in turn depends only on the Pearson's coefficient of the data. We discuss clustering algorithms that provide a fast and reliable approximation to maximum likelihood configurations. Compared to standard clustering methods, our approach has the advantages that (i) it is parameter free, (ii) the number of clusters need not be fixed in advance and (iii) the interpretation of the results is transparent. In order to test our approach and compare it with standard clustering algorithms, we analyze two very different data sets: time series of financial market returns and gene expression data. We find that different maximization algorithms produce similar cluster structures whereas the outcome of standard algorithms has a much wider variability.

  6. Clustering of resting state networks.

    Directory of Open Access Journals (Sweden)

    Megan H Lee

    Full Text Available The goal of the study was to demonstrate a hierarchical structure of resting state activity in the healthy brain using a data-driven clustering algorithm.The fuzzy-c-means clustering algorithm was applied to resting state fMRI data in cortical and subcortical gray matter from two groups acquired separately, one of 17 healthy individuals and the second of 21 healthy individuals. Different numbers of clusters and different starting conditions were used. A cluster dispersion measure determined the optimal numbers of clusters. An inner product metric provided a measure of similarity between different clusters. The two cluster result found the task-negative and task-positive systems. The cluster dispersion measure was minimized with seven and eleven clusters. Each of the clusters in the seven and eleven cluster result was associated with either the task-negative or task-positive system. Applying the algorithm to find seven clusters recovered previously described resting state networks, including the default mode network, frontoparietal control network, ventral and dorsal attention networks, somatomotor, visual, and language networks. The language and ventral attention networks had significant subcortical involvement. This parcellation was consistently found in a large majority of algorithm runs under different conditions and was robust to different methods of initialization.The clustering of resting state activity using different optimal numbers of clusters identified resting state networks comparable to previously obtained results. This work reinforces the observation that resting state networks are hierarchically organized.

  7. Unsupervised detection and removal of muscle artifacts from scalp EEG recordings using canonical correlation analysis, wavelets and random forests.

    Science.gov (United States)

    Anastasiadou, Maria N; Christodoulakis, Manolis; Papathanasiou, Eleftherios S; Papacostas, Savvas S; Mitsis, Georgios D

    2017-09-01

    This paper proposes supervised and unsupervised algorithms for automatic muscle artifact detection and removal from long-term EEG recordings, which combine canonical correlation analysis (CCA) and wavelets with random forests (RF). The proposed algorithms first perform CCA and continuous wavelet transform of the canonical components to generate a number of features which include component autocorrelation values and wavelet coefficient magnitude values. A subset of the most important features is subsequently selected using RF and labelled observations (supervised case) or synthetic data constructed from the original observations (unsupervised case). The proposed algorithms are evaluated using realistic simulation data as well as 30min epochs of non-invasive EEG recordings obtained from ten patients with epilepsy. We assessed the performance of the proposed algorithms using classification performance and goodness-of-fit values for noisy and noise-free signal windows. In the simulation study, where the ground truth was known, the proposed algorithms yielded almost perfect performance. In the case of experimental data, where expert marking was performed, the results suggest that both the supervised and unsupervised algorithm versions were able to remove artifacts without affecting noise-free channels considerably, outperforming standard CCA, independent component analysis (ICA) and Lagged Auto-Mutual Information Clustering (LAMIC). The proposed algorithms achieved excellent performance for both simulation and experimental data. Importantly, for the first time to our knowledge, we were able to perform entirely unsupervised artifact removal, i.e. without using already marked noisy data segments, achieving performance that is comparable to the supervised case. Overall, the results suggest that the proposed algorithms yield significant future potential for improving EEG signal quality in research or clinical settings without the need for marking by expert

  8. iHAT: interactive Hierarchical Aggregation Table for Genetic Association Data

    Directory of Open Access Journals (Sweden)

    Heinrich Julian

    2012-05-01

    Full Text Available Abstract In the search for single-nucleotide polymorphisms which influence the observable phenotype, genome wide association studies have become an important technique for the identification of associations between genotype and phenotype of a diverse set of sequence-based data. We present a methodology for the visual assessment of single-nucleotide polymorphisms using interactive hierarchical aggregation techniques combined with methods known from traditional sequence browsers and cluster heatmaps. Our tool, the interactive Hierarchical Aggregation Table (iHAT, facilitates the visualization of multiple sequence alignments, associated metadata, and hierarchical clusterings. Different color maps and aggregation strategies as well as filtering options support the user in finding correlations between sequences and metadata. Similar to other visualizations such as parallel coordinates or heatmaps, iHAT relies on the human pattern-recognition ability for spotting patterns that might indicate correlation or anticorrelation. We demonstrate iHAT using artificial and real-world datasets for DNA and protein association studies as well as expression Quantitative Trait Locus data.

  9. Glaucomatous patterns in Frequency Doubling Technology (FDT) perimetry data identified by unsupervised machine learning classifiers.

    Science.gov (United States)

    Bowd, Christopher; Weinreb, Robert N; Balasubramanian, Madhusudhanan; Lee, Intae; Jang, Giljin; Yousefi, Siamak; Zangwill, Linda M; Medeiros, Felipe A; Girkin, Christopher A; Liebmann, Jeffrey M; Goldbaum, Michael H

    2014-01-01

    The variational Bayesian independent component analysis-mixture model (VIM), an unsupervised machine-learning classifier, was used to automatically separate Matrix Frequency Doubling Technology (FDT) perimetry data into clusters of healthy and glaucomatous eyes, and to identify axes representing statistically independent patterns of defect in the glaucoma clusters. FDT measurements were obtained from 1,190 eyes with normal FDT results and 786 eyes with abnormal FDT results from the UCSD-based Diagnostic Innovations in Glaucoma Study (DIGS) and African Descent and Glaucoma Evaluation Study (ADAGES). For all eyes, VIM input was 52 threshold test points from the 24-2 test pattern, plus age. FDT mean deviation was -1.00 dB (S.D. = 2.80 dB) and -5.57 dB (S.D. = 5.09 dB) in FDT-normal eyes and FDT-abnormal eyes, respectively (p<0.001). VIM identified meaningful clusters of FDT data and positioned a set of statistically independent axes through the mean of each cluster. The optimal VIM model separated the FDT fields into 3 clusters. Cluster N contained primarily normal fields (1109/1190, specificity 93.1%) and clusters G1 and G2 combined, contained primarily abnormal fields (651/786, sensitivity 82.8%). For clusters G1 and G2 the optimal number of axes were 2 and 5, respectively. Patterns automatically generated along axes within the glaucoma clusters were similar to those known to be indicative of glaucoma. Fields located farther from the normal mean on each glaucoma axis showed increasing field defect severity. VIM successfully separated FDT fields from healthy and glaucoma eyes without a priori information about class membership, and identified familiar glaucomatous patterns of loss.

  10. Subspace K-means clustering.

    Science.gov (United States)

    Timmerman, Marieke E; Ceulemans, Eva; De Roover, Kim; Van Leeuwen, Karla

    2013-12-01

    To achieve an insightful clustering of multivariate data, we propose subspace K-means. Its central idea is to model the centroids and cluster residuals in reduced spaces, which allows for dealing with a wide range of cluster types and yields rich interpretations of the clusters. We review the existing related clustering methods, including deterministic, stochastic, and unsupervised learning approaches. To evaluate subspace K-means, we performed a comparative simulation study, in which we manipulated the overlap of subspaces, the between-cluster variance, and the error variance. The study shows that the subspace K-means algorithm is sensitive to local minima but that the problem can be reasonably dealt with by using partitions of various cluster procedures as a starting point for the algorithm. Subspace K-means performs very well in recovering the true clustering across all conditions considered and appears to be superior to its competitor methods: K-means, reduced K-means, factorial K-means, mixtures of factor analyzers (MFA), and MCLUST. The best competitor method, MFA, showed a performance similar to that of subspace K-means in easy conditions but deteriorated in more difficult ones. Using data from a study on parental behavior, we show that subspace K-means analysis provides a rich insight into the cluster characteristics, in terms of both the relative positions of the clusters (via the centroids) and the shape of the clusters (via the within-cluster residuals).

  11. Projection-based curve clustering

    International Nuclear Information System (INIS)

    Auder, Benjamin; Fischer, Aurelie

    2012-01-01

    This paper focuses on unsupervised curve classification in the context of nuclear industry. At the Commissariat a l'Energie Atomique (CEA), Cadarache (France), the thermal-hydraulic computer code CATHARE is used to study the reliability of reactor vessels. The code inputs are physical parameters and the outputs are time evolution curves of a few other physical quantities. As the CATHARE code is quite complex and CPU time-consuming, it has to be approximated by a regression model. This regression process involves a clustering step. In the present paper, the CATHARE output curves are clustered using a k-means scheme, with a projection onto a lower dimensional space. We study the properties of the empirically optimal cluster centres found by the clustering method based on projections, compared with the 'true' ones. The choice of the projection basis is discussed, and an algorithm is implemented to select the best projection basis among a library of orthonormal bases. The approach is illustrated on a simulated example and then applied to the industrial problem. (authors)

  12. A High Throughput Ambient Mass Spectrometric Approach to Species Identification and Classification from Chemical Fingerprint Signatures

    OpenAIRE

    Musah, Rabi A.; Espinoza, Edgard O.; Cody, Robert B.; Lesiak, Ashton D.; Christensen, Earl D.; Moore, Hannah E.; Maleknia, Simin; Drijfhout, Falko P.

    2015-01-01

    A high throughput method for species identification and classification through chemometric processing of direct analysis in real time (DART) mass spectrometry-derived fingerprint signatures has been developed. The method entails introduction of samples to the open air space between the DART ion source and the mass spectrometer inlet, with the entire observed mass spectral fingerprint subjected to unsupervised hierarchical clustering processing. A range of both polar and non-polar chemotypes a...

  13. ESPRIT-Forest: Parallel clustering of massive amplicon sequence data in subquadratic time.

    Science.gov (United States)

    Cai, Yunpeng; Zheng, Wei; Yao, Jin; Yang, Yujie; Mai, Volker; Mao, Qi; Sun, Yijun

    2017-04-01

    The rapid development of sequencing technology has led to an explosive accumulation of genomic sequence data. Clustering is often the first step to perform in sequence analysis, and hierarchical clustering is one of the most commonly used approaches for this purpose. However, it is currently computationally expensive to perform hierarchical clustering of extremely large sequence datasets due to its quadratic time and space complexities. In this paper we developed a new algorithm called ESPRIT-Forest for parallel hierarchical clustering of sequences. The algorithm achieves subquadratic time and space complexity and maintains a high clustering accuracy comparable to the standard method. The basic idea is to organize sequences into a pseudo-metric based partitioning tree for sub-linear time searching of nearest neighbors, and then use a new multiple-pair merging criterion to construct clusters in parallel using multiple threads. The new algorithm was tested on the human microbiome project (HMP) dataset, currently one of the largest published microbial 16S rRNA sequence dataset. Our experiment demonstrated that with the power of parallel computing it is now compu- tationally feasible to perform hierarchical clustering analysis of tens of millions of sequences. The software is available at http://www.acsu.buffalo.edu/∼yijunsun/lab/ESPRIT-Forest.html.

  14. [Cluster analysis in biomedical researches].

    Science.gov (United States)

    Akopov, A S; Moskovtsev, A A; Dolenko, S A; Savina, G D

    2013-01-01

    Cluster analysis is one of the most popular methods for the analysis of multi-parameter data. The cluster analysis reveals the internal structure of the data, group the separate observations on the degree of their similarity. The review provides a definition of the basic concepts of cluster analysis, and discusses the most popular clustering algorithms: k-means, hierarchical algorithms, Kohonen networks algorithms. Examples are the use of these algorithms in biomedical research.

  15. Musical genres: beating to the rhythms of different drums

    Science.gov (United States)

    Correa, Debora C.; Saito, Jose H.; Costa, Luciano da F.

    2010-05-01

    Online music databases have increased significantly as a consequence of the rapid growth of the Internet and digital audio, requiring the development of faster and more efficient tools for music content analysis. Musical genres are widely used to organize music collections. In this paper, the problem of automatic single and multi-label music genre classification is addressed by exploring rhythm-based features obtained from a respective complex network representation. A Markov model is built in order to analyse the temporal sequence of rhythmic notation events. Feature analysis is performed by using two multivariate statistical approaches: principal components analysis (unsupervised) and linear discriminant analysis (supervised). Similarly, two classifiers are applied in order to identify the category of rhythms: parametric Bayesian classifier under the Gaussian hypothesis (supervised) and agglomerative hierarchical clustering (unsupervised). Qualitative results obtained by using the kappa coefficient and the obtained clusters corroborated the effectiveness of the proposed method.

  16. Musical genres: beating to the rhythms of different drums

    Energy Technology Data Exchange (ETDEWEB)

    Correa, Debora C; Costa, Luciano da F [Instituto de Fisica de Sao Carlos - Universidade de Sao Paulo, Av. Trabalhador Sao Carlense 400, Caixa Postal 369, CEP 13560-970, Sao Carlos, Sao Paulo (Brazil); Saito, Jose H, E-mail: deboracorrea@ursa.ifsc.usp.b, E-mail: luciano@ursa.ifsc.usp.b [Departamento de Computacao-Universidade Federal de Sao Carlos, Rodovia Washington Luis, km 235, SP-310, CEP 13565-905, Sao Carlos, Sao Paulo (Brazil)

    2010-05-15

    Online music databases have increased significantly as a consequence of the rapid growth of the Internet and digital audio, requiring the development of faster and more efficient tools for music content analysis. Musical genres are widely used to organize music collections. In this paper, the problem of automatic single and multi-label music genre classification is addressed by exploring rhythm-based features obtained from a respective complex network representation. A Markov model is built in order to analyse the temporal sequence of rhythmic notation events. Feature analysis is performed by using two multivariate statistical approaches: principal components analysis (unsupervised) and linear discriminant analysis (supervised). Similarly, two classifiers are applied in order to identify the category of rhythms: parametric Bayesian classifier under the Gaussian hypothesis (supervised) and agglomerative hierarchical clustering (unsupervised). Qualitative results obtained by using the kappa coefficient and the obtained clusters corroborated the effectiveness of the proposed method.

  17. Musical genres: beating to the rhythms of different drums

    International Nuclear Information System (INIS)

    Correa, Debora C; Costa, Luciano da F; Saito, Jose H

    2010-01-01

    Online music databases have increased significantly as a consequence of the rapid growth of the Internet and digital audio, requiring the development of faster and more efficient tools for music content analysis. Musical genres are widely used to organize music collections. In this paper, the problem of automatic single and multi-label music genre classification is addressed by exploring rhythm-based features obtained from a respective complex network representation. A Markov model is built in order to analyse the temporal sequence of rhythmic notation events. Feature analysis is performed by using two multivariate statistical approaches: principal components analysis (unsupervised) and linear discriminant analysis (supervised). Similarly, two classifiers are applied in order to identify the category of rhythms: parametric Bayesian classifier under the Gaussian hypothesis (supervised) and agglomerative hierarchical clustering (unsupervised). Qualitative results obtained by using the kappa coefficient and the obtained clusters corroborated the effectiveness of the proposed method.

  18. Hierarchical MAS based control strategy for microgrid

    Energy Technology Data Exchange (ETDEWEB)

    Xiao, Z.; Li, T.; Huang, M.; Shi, J.; Yang, J.; Yu, J. [School of Information Science and Engineering, Yunnan University, Kunming 650091 (China); Xiao, Z. [School of Electrical and Electronic Engineering, Nanyang Technological University, Western Catchment Area, 639798 (Singapore); Wu, W. [Communication Branch of Yunnan Power Grid Corporation, Kunming, Yunnan 650217 (China)

    2010-09-15

    Microgrids have become a hot topic driven by the dual pressures of environmental protection concerns and the energy crisis. In this paper, a challenge for the distributed control of a modern electric grid incorporating clusters of residential microgrids is elaborated and a hierarchical multi-agent system (MAS) is proposed as a solution. The issues of how to realize the hierarchical MAS and how to improve coordination and control strategies are discussed. Based on MATLAB and ZEUS platforms, bilateral switching between grid-connected mode and island mode is performed under control of the proposed MAS to enhance and support its effectiveness. (authors)

  19. Hierarchical cluster analysis of technical replicates to identify interferents in untargeted mass spectrometry metabolomics.

    Science.gov (United States)

    Caesar, Lindsay K; Kvalheim, Olav M; Cech, Nadja B

    2018-08-27

    Mass spectral data sets often contain experimental artefacts, and data filtering prior to statistical analysis is crucial to extract reliable information. This is particularly true in untargeted metabolomics analyses, where the analyte(s) of interest are not known a priori. It is often assumed that chemical interferents (i.e. solvent contaminants such as plasticizers) are consistent across samples, and can be removed by background subtraction from blank injections. On the contrary, it is shown here that chemical contaminants may vary in abundance across each injection, potentially leading to their misidentification as relevant sample components. With this metabolomics study, we demonstrate the effectiveness of hierarchical cluster analysis (HCA) of replicate injections (technical replicates) as a methodology to identify chemical interferents and reduce their contaminating contribution to metabolomics models. Pools of metabolites with varying complexity were prepared from the botanical Angelica keiskei Koidzumi and spiked with known metabolites. Each set of pools was analyzed in triplicate and at multiple concentrations using ultraperformance liquid chromatography coupled to mass spectrometry (UPLC-MS). Before filtering, HCA failed to cluster replicates in the data sets. To identify contaminant peaks, we developed a filtering process that evaluated the relative peak area variance of each variable within triplicate injections. These interferent peaks were found across all samples, but did not show consistent peak area from injection to injection, even when evaluating the same chemical sample. This filtering process identified 128 ions that appear to originate from the UPLC-MS system. Data sets collected for a high number of pools with comparatively simple chemical composition were highly influenced by these chemical interferents, as were samples that were analyzed at a low concentration. When chemical interferent masses were removed, technical replicates clustered in

  20. The association between content of the elements S, Cl, K, Fe, Cu, Zn and Br in normal and cirrhotic liver tissue from Danes and Greenlandic Inuit examined by dual hierarchical clustering analysis.

    Science.gov (United States)

    Laursen, Jens; Milman, Nils; Pind, Niels; Pedersen, Henrik; Mulvad, Gert

    2014-01-01

    Meta-analysis of previous studies evaluating associations between content of elements sulphur (S), chlorine (Cl), potassium (K), iron (Fe), copper (Cu), zinc (Zn) and bromine (Br) in normal and cirrhotic autopsy liver tissue samples. Normal liver samples from 45 Greenlandic Inuit, median age 60 years and from 71 Danes, median age 61 years. Cirrhotic liver samples from 27 Danes, median age 71 years. Element content was measured using X-ray fluorescence spectrometry. Dual hierarchical clustering analysis, creating a dual dendrogram, one clustering element contents according to calculated similarities, one clustering elements according to correlation coefficients between the element contents, both using Euclidian distance and Ward Procedure. One dendrogram separated subjects in 7 clusters showing no differences in ethnicity, gender or age. The analysis discriminated between elements in normal and cirrhotic livers. The other dendrogram clustered elements in four clusters: sulphur and chlorine; copper and bromine; potassium and zinc; iron. There were significant correlations between the elements in normal liver samples: S was associated with Cl, K, Br and Zn; Cl with S and Br; K with S, Br and Zn; Cu with Br. Zn with S and K. Br with S, Cl, K and Cu. Fe did not show significant associations with any other element. In contrast to simple statistical methods, which analyses content of elements separately one by one, dual hierarchical clustering analysis incorporates all elements at the same time and can be used to examine the linkage and interplay between multiple elements in tissue samples. Copyright © 2013 Elsevier GmbH. All rights reserved.

  1. Mastication Evaluation With Unsupervised Learning: Using an Inertial Sensor-Based System

    Science.gov (United States)

    Lucena, Caroline Vieira; Lacerda, Marcelo; Caldas, Rafael; De Lima Neto, Fernando Buarque

    2018-01-01

    There is a direct relationship between the prevalence of musculoskeletal disorders of the temporomandibular joint and orofacial disorders. A well-elaborated analysis of the jaw movements provides relevant information for healthcare professionals to conclude their diagnosis. Different approaches have been explored to track jaw movements such that the mastication analysis is getting less subjective; however, all methods are still highly subjective, and the quality of the assessments depends much on the experience of the health professional. In this paper, an accurate and non-invasive method based on a commercial low-cost inertial sensor (MPU6050) to measure jaw movements is proposed. The jaw-movement feature values are compared to the obtained with clinical analysis, showing no statistically significant difference between both methods. Moreover, We propose to use unsupervised paradigm approaches to cluster mastication patterns of healthy subjects and simulated patients with facial trauma. Two techniques were used in this paper to instantiate the method: Kohonen’s Self-Organizing Maps and K-Means Clustering. Both algorithms have excellent performances to process jaw-movements data, showing encouraging results and potential to bring a full assessment of the masticatory function. The proposed method can be applied in real-time providing relevant dynamic information for health-care professionals. PMID:29651365

  2. Mastication Evaluation With Unsupervised Learning: Using an Inertial Sensor-Based System.

    Science.gov (United States)

    Lucena, Caroline Vieira; Lacerda, Marcelo; Caldas, Rafael; De Lima Neto, Fernando Buarque; Rativa, Diego

    2018-01-01

    There is a direct relationship between the prevalence of musculoskeletal disorders of the temporomandibular joint and orofacial disorders. A well-elaborated analysis of the jaw movements provides relevant information for healthcare professionals to conclude their diagnosis. Different approaches have been explored to track jaw movements such that the mastication analysis is getting less subjective; however, all methods are still highly subjective, and the quality of the assessments depends much on the experience of the health professional. In this paper, an accurate and non-invasive method based on a commercial low-cost inertial sensor (MPU6050) to measure jaw movements is proposed. The jaw-movement feature values are compared to the obtained with clinical analysis, showing no statistically significant difference between both methods. Moreover, We propose to use unsupervised paradigm approaches to cluster mastication patterns of healthy subjects and simulated patients with facial trauma. Two techniques were used in this paper to instantiate the method: Kohonen's Self-Organizing Maps and K-Means Clustering. Both algorithms have excellent performances to process jaw-movements data, showing encouraging results and potential to bring a full assessment of the masticatory function. The proposed method can be applied in real-time providing relevant dynamic information for health-care professionals.

  3. Unsupervised versus Supervised Identification of Prognostic Factors in Patients with Localized Retroperitoneal Sarcoma: A Data Clustering and Mahalanobis Distance Approach

    Directory of Open Access Journals (Sweden)

    Rita De Sanctis

    2018-01-01

    Full Text Available The aim of this report is to unveil specific prognostic factors for retroperitoneal sarcoma (RPS patients by univariate and multivariate statistical techniques. A phase I-II study on localized RPS treated with high-dose ifosfamide and radiotherapy followed by surgery (ISG-STS 0303 protocol demonstrated that chemo/radiotherapy was safe and increased the 3-year relapse-free survival (RFS with respect to historical controls. Of 70 patients, twenty-six developed local, 10 distant, and 5 combined relapse. Median disease-free interval (DFI was 29.47 months. According to a discriminant function analysis, DFI, histology, relapse pattern, and the first treatment approach at relapse had a statistically significant prognostic impact. Based on scientific literature and clinical expertise, clinicopathological data were analyzed using both a supervised and an unsupervised classification method to predict the prognosis, with similar sample sizes (66 and 65, resp., in casewise approach and 70 in mean-substitution one. This is the first attempt to predict patients’ prognosis by means of multivariate statistics, and in this light, it looks noticable that (i some clinical data have a well-defined prognostic value, (ii the unsupervised model produced comparable results with respect to the supervised one, and (iii the appropriate combination of both models appears fruitful and easily extensible to different clinical contexts.

  4. A Trajectory Regression Clustering Technique Combining a Novel Fuzzy C-Means Clustering Algorithm with the Least Squares Method

    Directory of Open Access Journals (Sweden)

    Xiangbing Zhou

    2018-04-01

    Full Text Available Rapidly growing GPS (Global Positioning System trajectories hide much valuable information, such as city road planning, urban travel demand, and population migration. In order to mine the hidden information and to capture better clustering results, a trajectory regression clustering method (an unsupervised trajectory clustering method is proposed to reduce local information loss of the trajectory and to avoid getting stuck in the local optimum. Using this method, we first define our new concept of trajectory clustering and construct a novel partitioning (angle-based partitioning method of line segments; second, the Lagrange-based method and Hausdorff-based K-means++ are integrated in fuzzy C-means (FCM clustering, which are used to maintain the stability and the robustness of the clustering process; finally, least squares regression model is employed to achieve regression clustering of the trajectory. In our experiment, the performance and effectiveness of our method is validated against real-world taxi GPS data. When comparing our clustering algorithm with the partition-based clustering algorithms (K-means, K-median, and FCM, our experimental results demonstrate that the presented method is more effective and generates a more reasonable trajectory.

  5. Hierarchical sets: analyzing pangenome structure through scalable set visualizations

    Science.gov (United States)

    2017-01-01

    Abstract Motivation: The increase in available microbial genome sequences has resulted in an increase in the size of the pangenomes being analyzed. Current pangenome visualizations are not intended for the pangenome sizes possible today and new approaches are necessary in order to convert the increase in available information to increase in knowledge. As the pangenome data structure is essentially a collection of sets we explore the potential for scalable set visualization as a tool for pangenome analysis. Results: We present a new hierarchical clustering algorithm based on set arithmetics that optimizes the intersection sizes along the branches. The intersection and union sizes along the hierarchy are visualized using a composite dendrogram and icicle plot, which, in pangenome context, shows the evolution of pangenome and core size along the evolutionary hierarchy. Outlying elements, i.e. elements whose presence pattern do not correspond with the hierarchy, can be visualized using hierarchical edge bundles. When applied to pangenome data this plot shows putative horizontal gene transfers between the genomes and can highlight relationships between genomes that is not represented by the hierarchy. We illustrate the utility of hierarchical sets by applying it to a pangenome based on 113 Escherichia and Shigella genomes and find it provides a powerful addition to pangenome analysis. Availability and Implementation: The described clustering algorithm and visualizations are implemented in the hierarchicalSets R package available from CRAN (https://cran.r-project.org/web/packages/hierarchicalSets) Contact: thomasp85@gmail.com Supplementary information: Supplementary data are available at Bioinformatics online. PMID:28130242

  6. Revisiting the variation of clustering coefficient of biological networks suggests new modular structure.

    Science.gov (United States)

    Hao, Dapeng; Ren, Cong; Li, Chuanxing

    2012-05-01

    A central idea in biology is the hierarchical organization of cellular processes. A commonly used method to identify the hierarchical modular organization of network relies on detecting a global signature known as variation of clustering coefficient (so-called modularity scaling). Although several studies have suggested other possible origins of this signature, it is still widely used nowadays to identify hierarchical modularity, especially in the analysis of biological networks. Therefore, a further and systematical investigation of this signature for different types of biological networks is necessary. We analyzed a variety of biological networks and found that the commonly used signature of hierarchical modularity is actually the reflection of spoke-like topology, suggesting a different view of network architecture. We proved that the existence of super-hubs is the origin that the clustering coefficient of a node follows a particular scaling law with degree k in metabolic networks. To study the modularity of biological networks, we systematically investigated the relationship between repulsion of hubs and variation of clustering coefficient. We provided direct evidences for repulsion between hubs being the underlying origin of the variation of clustering coefficient, and found that for biological networks having no anti-correlation between hubs, such as gene co-expression network, the clustering coefficient doesn't show dependence of degree. Here we have shown that the variation of clustering coefficient is neither sufficient nor exclusive for a network to be hierarchical. Our results suggest the existence of spoke-like modules as opposed to "deterministic model" of hierarchical modularity, and suggest the need to reconsider the organizational principle of biological hierarchy.

  7. Robust multi-scale clustering of large DNA microarray datasets with the consensus algorithm

    DEFF Research Database (Denmark)

    Grotkjær, Thomas; Winther, Ole; Regenberg, Birgitte

    2006-01-01

    Motivation: Hierarchical and relocation clustering (e.g. K-means and self-organizing maps) have been successful tools in the display and analysis of whole genome DNA microarray expression data. However, the results of hierarchical clustering are sensitive to outliers, and most relocation methods...... analysis by collecting re-occurring clustering patterns in a co-occurrence matrix. The results show that consensus clustering obtained from clustering multiple times with Variational Bayes Mixtures of Gaussians or K-means significantly reduces the classification error rate for a simulated dataset...

  8. D Nearest Neighbour Search Using a Clustered Hierarchical Tree Structure

    Science.gov (United States)

    Suhaibah, A.; Uznir, U.; Anton, F.; Mioc, D.; Rahman, A. A.

    2016-06-01

    Locating and analysing the location of new stores or outlets is one of the common issues facing retailers and franchisers. This is due to assure that new opening stores are at their strategic location to attract the highest possible number of customers. Spatial information is used to manage, maintain and analyse these store locations. However, since the business of franchising and chain stores in urban areas runs within high rise multi-level buildings, a three-dimensional (3D) method is prominently required in order to locate and identify the surrounding information such as at which level of the franchise unit will be located or is the franchise unit located is at the best level for visibility purposes. One of the common used analyses used for retrieving the surrounding information is Nearest Neighbour (NN) analysis. It uses a point location and identifies the surrounding neighbours. However, with the immense number of urban datasets, the retrieval and analysis of nearest neighbour information and their efficiency will become more complex and crucial. In this paper, we present a technique to retrieve nearest neighbour information in 3D space using a clustered hierarchical tree structure. Based on our findings, the proposed approach substantially showed an improvement of response time analysis compared to existing approaches of spatial access methods in databases. The query performance was tested using a dataset consisting of 500,000 point locations building and franchising unit. The results are presented in this paper. Another advantage of this structure is that it also offers a minimal overlap and coverage among nodes which can reduce repetitive data entry.

  9. Unsupervised Anomaly Detection for Liquid-Fueled Rocket Prop...

    Data.gov (United States)

    National Aeronautics and Space Administration — Title: Unsupervised Anomaly Detection for Liquid-Fueled Rocket Propulsion Health Monitoring. Abstract: This article describes the results of applying four...

  10. Hierarchical silica particles by dynamic multicomponent assembly

    DEFF Research Database (Denmark)

    Wu, Z. W.; Hu, Q. Y.; Pang, J. B.

    2005-01-01

    Abstract: Aerosol-assisted assembly of mesoporous silica particles with hierarchically controllable pore structure has been prepared using cetyltrimethylammonium bromide (CTAB) and poly(propylene oxide) (PPO, H[OCH(CH3)CH2],OH) as co-templates. Addition of the hydrophobic PPO significantly...... influences the delicate hydrophilic-hydrophobic balance in the well-studied CTAB-silicate co-assembling system, resulting in various mesostructures (such as hexagonal, lamellar, and hierarchical structure). The co-assembly of CTAB, silicate clusters, and a low-molecular-weight PPO (average M-n 425) results...... in a uniform lamellar structure, while the use of a high-molecular-weight PPO (average M-n 2000), which is more hydrophobic, leads to the formation of hierarchical pore structure that contains meso-meso or meso-macro pore structure. The role of PPO additives on the mesostructure evolution in the CTAB...

  11. A local adaptive algorithm for emerging scale-free hierarchical networks

    International Nuclear Information System (INIS)

    Gomez Portillo, I J; Gleiser, P M

    2010-01-01

    In this work we study a growing network model with chaotic dynamical units that evolves using a local adaptive rewiring algorithm. Using numerical simulations we show that the model allows for the emergence of hierarchical networks. First, we show that the networks that emerge with the algorithm present a wide degree distribution that can be fitted by a power law function, and thus are scale-free networks. Using the LaNet-vi visualization tool we present a graphical representation that reveals a central core formed only by hubs, and also show the presence of a preferential attachment mechanism. In order to present a quantitative analysis of the hierarchical structure we analyze the clustering coefficient. In particular, we show that as the network grows the clustering becomes independent of system size, and also presents a power law decay as a function of the degree. Finally, we compare our results with a similar version of the model that has continuous non-linear phase oscillators as dynamical units. The results show that local interactions play a fundamental role in the emergence of hierarchical networks.

  12. Random clustering ferns for multimodal object recognition

    OpenAIRE

    Villamizar Vergel, Michael Alejandro; Garrell Zulueta, Anais; Sanfeliu Cortés, Alberto; Moreno-Noguer, Francesc

    2017-01-01

    The final publication is available at link.springer.com We propose an efficient and robust method for the recognition of objects exhibiting multiple intra-class modes, where each one is associated with a particular object appearance. The proposed method, called random clustering ferns, combines synergically a single and real-time classifier, based on the boosted assembling of extremely randomized trees (ferns), with an unsupervised and probabilistic approach in order to recognize efficient...

  13. Maximum Margin Clustering of Hyperspectral Data

    Science.gov (United States)

    Niazmardi, S.; Safari, A.; Homayouni, S.

    2013-09-01

    In recent decades, large margin methods such as Support Vector Machines (SVMs) are supposed to be the state-of-the-art of supervised learning methods for classification of hyperspectral data. However, the results of these algorithms mainly depend on the quality and quantity of available training data. To tackle down the problems associated with the training data, the researcher put effort into extending the capability of large margin algorithms for unsupervised learning. One of the recent proposed algorithms is Maximum Margin Clustering (MMC). The MMC is an unsupervised SVMs algorithm that simultaneously estimates both the labels and the hyperplane parameters. Nevertheless, the optimization of the MMC algorithm is a non-convex problem. Most of the existing MMC methods rely on the reformulating and the relaxing of the non-convex optimization problem as semi-definite programs (SDP), which are computationally very expensive and only can handle small data sets. Moreover, most of these algorithms are two-class classification, which cannot be used for classification of remotely sensed data. In this paper, a new MMC algorithm is used that solve the original non-convex problem using Alternative Optimization method. This algorithm is also extended for multi-class classification and its performance is evaluated. The results of the proposed algorithm show that the algorithm has acceptable results for hyperspectral data clustering.

  14. Ultrathin mesoporous Co_3O_4 nanosheets-constructed hierarchical clusters as high rate capability and long life anode materials for lithium-ion batteries

    International Nuclear Information System (INIS)

    Wu, Shengming; Xia, Tian; Wang, Jingping; Lu, Feifei; Xu, Chunbo; Zhang, Xianfa; Huo, Lihua; Zhao, Hui

    2017-01-01

    Graphical abstract: Ultrathin mesoporous Co_3O_4 nanosheets-constructed hierarchical clusters (UMCN-HCs) have been successfully synthesized via a facile hydrothermal method followed by a subsequent thermolysis treatment. When tested as anode materials for LIBs, UMCN-HCs achieve high reversible capacity, good long cycling life, and rate capability. - Highlights: • UMCN-HCs show high capacity, excellent stability, and good rate capability. • UMCN-HCs retain a capacity of 1067 mAh g"−"1 after 100 cycles at 100 mA g"−"1. • UMCN-HCs deliver a capacity of 507 mAh g"−"1 after 500 cycles at 2 A g"−"1. - Abstract: Herein, Ultrathin mesoporous Co_3O_4 nanosheets-constructed hierarchical clusters (UMCN-HCs) have been successfully synthesized via a facile hydrothermal method followed by a subsequent thermolysis treatment at 600 °C in air. The products consist of cluster-like Co_3O_4 microarchitectures, which are assembled by numerous ultrathin mesoporous Co_3O_4 nanosheets. When tested as anode materials for lithium-ion batteries, UMCN-HCs deliver a high reversible capacity of 1067 mAh g"−"1 at a current density of 100 mA g"−"1 after 100 cycles. Even at 2 A g"−"1, a stable capacity as high as 507 mAh g"−"1 can be achieved after 500 cycles. The high reversible capacity, excellent cycling stability, and good rate capability of UMCN-HCs may be attributed to their mesoporous sheet-like nanostructure. The sheet-layered structure of UMCN-HCs may buffer the volume change during the lithiation-delithiation process, and the mesoporous characteristic make lithium-ion transfer more easily at the interface between the active electrode and the electrolyte.

  15. Unsupervised process monitoring and fault diagnosis with machine learning methods

    CERN Document Server

    Aldrich, Chris

    2013-01-01

    This unique text/reference describes in detail the latest advances in unsupervised process monitoring and fault diagnosis with machine learning methods. Abundant case studies throughout the text demonstrate the efficacy of each method in real-world settings. The broad coverage examines such cutting-edge topics as the use of information theory to enhance unsupervised learning in tree-based methods, the extension of kernel methods to multiple kernel learning for feature extraction from data, and the incremental training of multilayer perceptrons to construct deep architectures for enhanced data

  16. Revisiting the variation of clustering coefficient of biological networks suggests new modular structure

    Directory of Open Access Journals (Sweden)

    Hao Dapeng

    2012-05-01

    Full Text Available Abstract Background A central idea in biology is the hierarchical organization of cellular processes. A commonly used method to identify the hierarchical modular organization of network relies on detecting a global signature known as variation of clustering coefficient (so-called modularity scaling. Although several studies have suggested other possible origins of this signature, it is still widely used nowadays to identify hierarchical modularity, especially in the analysis of biological networks. Therefore, a further and systematical investigation of this signature for different types of biological networks is necessary. Results We analyzed a variety of biological networks and found that the commonly used signature of hierarchical modularity is actually the reflection of spoke-like topology, suggesting a different view of network architecture. We proved that the existence of super-hubs is the origin that the clustering coefficient of a node follows a particular scaling law with degree k in metabolic networks. To study the modularity of biological networks, we systematically investigated the relationship between repulsion of hubs and variation of clustering coefficient. We provided direct evidences for repulsion between hubs being the underlying origin of the variation of clustering coefficient, and found that for biological networks having no anti-correlation between hubs, such as gene co-expression network, the clustering coefficient doesn’t show dependence of degree. Conclusions Here we have shown that the variation of clustering coefficient is neither sufficient nor exclusive for a network to be hierarchical. Our results suggest the existence of spoke-like modules as opposed to “deterministic model” of hierarchical modularity, and suggest the need to reconsider the organizational principle of biological hierarchy.

  17. Performance quantification of clustering algorithms for false positive removal in fMRI by ROC curves

    Directory of Open Access Journals (Sweden)

    André Salles Cunha Peres

    Full Text Available Abstract Introduction Functional magnetic resonance imaging (fMRI is a non-invasive technique that allows the detection of specific cerebral functions in humans based on hemodynamic changes. The contrast changes are about 5%, making visual inspection impossible. Thus, statistic strategies are applied to infer which brain region is engaged in a task. However, the traditional methods like general linear model and cross-correlation utilize voxel-wise calculation, introducing a lot of false-positive data. So, in this work we tested post-processing cluster algorithms to diminish the false-positives. Methods In this study, three clustering algorithms (the hierarchical cluster, k-means and self-organizing maps were tested and compared for false-positive removal in the post-processing of cross-correlation analyses. Results Our results showed that the hierarchical cluster presented the best performance to remove the false positives in fMRI, being 2.3 times more accurate than k-means, and 1.9 times more accurate than self-organizing maps. Conclusion The hierarchical cluster presented the best performance in false-positive removal because it uses the inconsistency coefficient threshold, while k-means and self-organizing maps utilize a priori cluster number (centroids and neurons number; thus, the hierarchical cluster avoids clustering scattered voxels, as the inconsistency coefficient threshold allows only the voxels to be clustered that are at a minimum distance to some cluster.

  18. A Simple Hierarchical Pooling Data Structure for Loop Closure

    Science.gov (United States)

    2016-10-16

    performance empirically on the KITTI [9], Oxford [6] and TUM RGB- D [29] datasets, as well as demonstrate extensions to general image retrieval on the...of a BoW where each word is an element of a dictionary of descriptors obtained off-line by hierarchical k-means clustering, with each word weighted by...to the inverse docu- ment frequency. This standard pipeline, with different clustering procedures to generate the dictionary and different features

  19. Spectrum Hole Identification in IEEE 802.22 WRAN using Unsupervised Learning

    Directory of Open Access Journals (Sweden)

    V. Balaji

    2016-01-01

    Full Text Available In this paper we present a Cooperative Spectrum Sensing (CSS algorithm for Cognitive Radios (CR based on IEEE 802.22Wireless Regional Area Network (WRAN standard. The core objective is to improve cooperative sensing efficiency which specifies how fast a decision can be reached in each round of cooperation (iteration to sense an appropriate number of channels/bands (i.e. 86 channels of 7MHz bandwidth as per IEEE 802.22 within a time constraint (channel sensing time. To meet this objective, we have developed CSS algorithm using unsupervised K-means clustering classification approach. The received energy level of each Secondary User (SU is considered as the parameter for determining channel availability. The performance of proposed algorithm is quantified in terms of detection accuracy, training and classification delay time. Further, the detection accuracy of our proposed scheme meets the requirement of IEEE 802.22 WRAN with the target probability of falsealrm as 0.1. All the simulations are carried out using Matlab tool.

  20. Hierarchical structure of the European countries based on debts as a percentage of GDP during the 2000-2011 period

    Science.gov (United States)

    Kantar, Ersin; Deviren, Bayram; Keskin, Mustafa

    2014-11-01

    We investigate hierarchical structures of the European countries by using debt as a percentage of Gross Domestic Product (GDP) of the countries as they change over a certain period of time. We obtain the topological properties among the countries based on debt as a percentage of GDP of European countries over the period 2000-2011 by using the concept of hierarchical structure methods (minimal spanning tree, (MST) and hierarchical tree, (HT)). This period is also divided into two sub-periods related to 2004 enlargement of the European Union, namely 2000-2004 and 2005-2011, in order to test various time-window and observe the temporal evolution. The bootstrap techniques is applied to see a value of statistical reliability of the links of the MSTs and HTs. The clustering linkage procedure is also used to observe the cluster structure more clearly. From the structural topologies of these trees, we identify different clusters of countries according to their level of debts. Our results show that by the debt crisis, the less and most affected Eurozone’s economies are formed as a cluster with each other in the MSTs and hierarchical trees.

  1. Quantum annealing for combinatorial clustering

    Science.gov (United States)

    Kumar, Vaibhaw; Bass, Gideon; Tomlin, Casey; Dulny, Joseph

    2018-02-01

    Clustering is a powerful machine learning technique that groups "similar" data points based on their characteristics. Many clustering algorithms work by approximating the minimization of an objective function, namely the sum of within-the-cluster distances between points. The straightforward approach involves examining all the possible assignments of points to each of the clusters. This approach guarantees the solution will be a global minimum; however, the number of possible assignments scales quickly with the number of data points and becomes computationally intractable even for very small datasets. In order to circumvent this issue, cost function minima are found using popular local search-based heuristic approaches such as k-means and hierarchical clustering. Due to their greedy nature, such techniques do not guarantee that a global minimum will be found and can lead to sub-optimal clustering assignments. Other classes of global search-based techniques, such as simulated annealing, tabu search, and genetic algorithms, may offer better quality results but can be too time-consuming to implement. In this work, we describe how quantum annealing can be used to carry out clustering. We map the clustering objective to a quadratic binary optimization problem and discuss two clustering algorithms which are then implemented on commercially available quantum annealing hardware, as well as on a purely classical solver "qbsolv." The first algorithm assigns N data points to K clusters, and the second one can be used to perform binary clustering in a hierarchical manner. We present our results in the form of benchmarks against well-known k-means clustering and discuss the advantages and disadvantages of the proposed techniques.

  2. Clustering performance comparison using K-means and expectation maximization algorithms.

    Science.gov (United States)

    Jung, Yong Gyu; Kang, Min Soo; Heo, Jun

    2014-11-14

    Clustering is an important means of data mining based on separating data categories by similar features. Unlike the classification algorithm, clustering belongs to the unsupervised type of algorithms. Two representatives of the clustering algorithms are the K -means and the expectation maximization (EM) algorithm. Linear regression analysis was extended to the category-type dependent variable, while logistic regression was achieved using a linear combination of independent variables. To predict the possibility of occurrence of an event, a statistical approach is used. However, the classification of all data by means of logistic regression analysis cannot guarantee the accuracy of the results. In this paper, the logistic regression analysis is applied to EM clusters and the K -means clustering method for quality assessment of red wine, and a method is proposed for ensuring the accuracy of the classification results.

  3. Specialization processes in on-line unsupervised learning

    NARCIS (Netherlands)

    Biehl, M.; Freking, A.; Reents, G.; Schlösser, E.

    1998-01-01

    From the recent analysis of supervised learning by on-line gradient descent in multilayered neural networks it is known that the necessary process of student specialization can be delayed significantly. We demonstrate that this phenomenon also occurs in various models of unsupervised learning. A

  4. Unsupervised Document Embedding With CNNs

    OpenAIRE

    Liu, Chundi; Zhao, Shunan; Volkovs, Maksims

    2017-01-01

    We propose a new model for unsupervised document embedding. Leading existing approaches either require complex inference or use recurrent neural networks (RNN) that are difficult to parallelize. We take a different route and develop a convolutional neural network (CNN) embedding model. Our CNN architecture is fully parallelizable resulting in over 10x speedup in inference time over RNN models. Parallelizable architecture enables to train deeper models where each successive layer has increasin...

  5. Gastric cancer differentiation using Fourier transform near-infrared spectroscopy with unsupervised pattern recognition

    Science.gov (United States)

    Yi, Wei-song; Cui, Dian-sheng; Li, Zhi; Wu, Lan-lan; Shen, Ai-guo; Hu, Ji-ming

    2013-01-01

    The manuscript has investigated the application of near-infrared (NIR) spectroscopy for differentiation gastric cancer. The 90 spectra from cancerous and normal tissues were collected from a total of 30 surgical specimens using Fourier transform near-infrared spectroscopy (FT-NIR) equipped with a fiber-optic probe. Major spectral differences were observed in the CH-stretching second overtone (9000-7000 cm-1), CH-stretching first overtone (6000-5200 cm-1), and CH-stretching combination (4500-4000 cm-1) regions. By use of unsupervised pattern recognition, such as principal component analysis (PCA) and cluster analysis (CA), all spectra were classified into cancerous and normal tissue groups with accuracy up to 81.1%. The sensitivity and specificity was 100% and 68.2%, respectively. These present results indicate that CH-stretching first, combination band and second overtone regions can serve as diagnostic markers for gastric cancer.

  6. Statistical measures of galaxy clustering

    International Nuclear Information System (INIS)

    Porter, D.H.

    1988-01-01

    Consideration is given to the large-scale distribution of galaxies and ways in which this distribution may be statistically measured. Galaxy clustering is hierarchical in nature, so that the positions of clusters of galaxies are themselves spatially clustered. A simple identification of groups of galaxies would be an inadequate description of the true richness of galaxy clustering. Current observations of the large-scale structure of the universe and modern theories of cosmology may be studied with a statistical description of the spatial and velocity distributions of galaxies. 8 refs

  7. Combining Unsupervised and Supervised Statistical Learning Methods for Currency Exchange Rate Forecasting

    OpenAIRE

    Vasiljeva, Polina

    2016-01-01

    In this thesis we revisit the challenging problem of forecasting currency exchange rate. We combine machine learning methods such as agglomerative hierarchical clustering and random forest to construct a two-step approach for predicting movements in currency exchange prices of the Swedish krona and the US dollar. We use a data set with over 200 predictors comprised of different financial and macro-economic time series and their transformations. We perform forecasting for one week ahead with d...

  8. Unsupervised learning of facial emotion decoding skills.

    Science.gov (United States)

    Huelle, Jan O; Sack, Benjamin; Broer, Katja; Komlewa, Irina; Anders, Silke

    2014-01-01

    Research on the mechanisms underlying human facial emotion recognition has long focussed on genetically determined neural algorithms and often neglected the question of how these algorithms might be tuned by social learning. Here we show that facial emotion decoding skills can be significantly and sustainably improved by practice without an external teaching signal. Participants saw video clips of dynamic facial expressions of five different women and were asked to decide which of four possible emotions (anger, disgust, fear, and sadness) was shown in each clip. Although no external information about the correctness of the participant's response or the sender's true affective state was provided, participants showed a significant increase of facial emotion recognition accuracy both within and across two training sessions two days to several weeks apart. We discuss several similarities and differences between the unsupervised improvement of facial decoding skills observed in the current study, unsupervised perceptual learning of simple stimuli described in previous studies and practice effects often observed in cognitive tasks.

  9. Unsupervised classification of neocortical activity patterns in neonatal and pre-juvenile rodents

    Directory of Open Access Journals (Sweden)

    Nicole eCichon

    2014-05-01

    Full Text Available Flexible communication within the brain, which relies on oscillatory activity, is not confined to adult neuronal networks. Experimental evidence has documented the presence of discontinuous patterns of oscillatory activity already during early development. Their highly variable spatial and time-frequency organization has been related to region specificity. However, it might be equally due to the absence of unitary criteria for classifying the early activity patterns, since they have been mainly characterized by visual inspection. Therefore, robust and unbiased methods for categorizing these discontinuous oscillations are needed for increasingly complex data sets from different labs. Here, we introduce an unsupervised detection and classification algorithm for the discontinuous activity patterns of rodents during early development. For this, firstly time windows with discontinuous oscillations vs. epochs of network silence were identified. In a second step, the major features of detected events were identified and processed by principal component analysis for deciding on their contribution to the classification of different oscillatory patterns. Finally, these patterns were categorized using an unsupervised cluster algorithm. The results were validated on manually characterized neonatal spindle bursts, which ubiquitously entrain neocortical areas of rats and mice, and prelimbic nested gamma spindle bursts. Moreover, the algorithm led to satisfactory results for oscillatory events that, due to increased similarity of their features, were more difficult to classify, e.g. during the pre-juvenile developmental period. Based on a linear classification, the optimal number of features to consider increased with the difficulty of detection. This algorithm allows the comparison of neonatal and pre-juvenile oscillatory patterns in their spatial and temporal organization. It might represent a first step for the unbiased elucidation of activity patterns

  10. Cities and regions in Britain through hierarchical percolation

    Science.gov (United States)

    Arcaute, Elsa; Molinero, Carlos; Hatna, Erez; Murcio, Roberto; Vargas-Ruiz, Camilo; Masucci, A. Paolo; Batty, Michael

    2016-04-01

    Urban systems present hierarchical structures at many different scales. These are observed as administrative regional delimitations which are the outcome of complex geographical, political and historical processes which leave almost indelible footprints on infrastructure such as the street network. In this work, we uncover a set of hierarchies in Britain at different scales using percolation theory on the street network and on its intersections which are the primary points of interaction and urban agglomeration. At the larger scales, the observed hierarchical structures can be interpreted as regional fractures of Britain, observed in various forms, from natural boundaries, such as National Parks, to regional divisions based on social class and wealth such as the well-known North-South divide. At smaller scales, cities are generated through recursive percolations on each of the emerging regional clusters. We examine the evolution of the morphology of the system as a whole, by measuring the fractal dimension of the clusters at each distance threshold in the percolation. We observe that this reaches a maximum plateau at a specific distance. The clusters defined at this distance threshold are in excellent correspondence with the boundaries of cities recovered from satellite images, and from previous methods using population density.

  11. Complexity of major UK companies between 2006 and 2010: Hierarchical structure method approach

    Science.gov (United States)

    Ulusoy, Tolga; Keskin, Mustafa; Shirvani, Ayoub; Deviren, Bayram; Kantar, Ersin; Çaǧrı Dönmez, Cem

    2012-11-01

    This study reports on topology of the top 40 UK companies that have been analysed for predictive verification of markets for the period 2006-2010, applying the concept of minimal spanning tree and hierarchical tree (HT) analysis. Construction of the minimal spanning tree (MST) and the hierarchical tree (HT) is confined to a brief description of the methodology and a definition of the correlation function between a pair of companies based on the London Stock Exchange (LSE) index in order to quantify synchronization between the companies. A derivation of hierarchical organization and the construction of minimal-spanning and hierarchical trees for the 2006-2008 and 2008-2010 periods have been used and the results validate the predictive verification of applied semantics. The trees are known as useful tools to perceive and detect the global structure, taxonomy and hierarchy in financial data. From these trees, two different clusters of companies in 2006 were detected. They also show three clusters in 2008 and two between 2008 and 2010, according to their proximity. The clusters match each other as regards their common production activities or their strong interrelationship. The key companies are generally given by major economic activities as expected. This work gives a comparative approach between MST and HT methods from statistical physics and information theory with analysis of financial markets that may give new valuable and useful information of the financial market dynamics.

  12. Clustering User Behavior in Scientific Collections

    OpenAIRE

    Blixhavn, Øystein Hoel

    2014-01-01

    This master thesis looks at how clustering techniques can be appliedto a collection of scientific documents. Approximately one year of serverlogs from the CERN Document Server (CDS) are analyzed and preprocessed.Based on the findings of this analysis, and a review of thecurrent state of the art, three different clustering methods are selectedfor further work: Simple k-Means, Hierarchical Agglomerative Clustering(HAC) and Graph Partitioning. In addition, a custom, agglomerativeclustering algor...

  13. Tractography segmentation using a hierarchical Dirichlet processes mixture model.

    Science.gov (United States)

    Wang, Xiaogang; Grimson, W Eric L; Westin, Carl-Fredrik

    2011-01-01

    In this paper, we propose a new nonparametric Bayesian framework to cluster white matter fiber tracts into bundles using a hierarchical Dirichlet processes mixture (HDPM) model. The number of clusters is automatically learned driven by data with a Dirichlet process (DP) prior instead of being manually specified. After the models of bundles have been learned from training data without supervision, they can be used as priors to cluster/classify fibers of new subjects for comparison across subjects. When clustering fibers of new subjects, new clusters can be created for structures not observed in the training data. Our approach does not require computing pairwise distances between fibers and can cluster a huge set of fibers across multiple subjects. We present results on several data sets, the largest of which has more than 120,000 fibers. Copyright © 2010 Elsevier Inc. All rights reserved.

  14. A new hierarchical method to find community structure in networks

    Science.gov (United States)

    Saoud, Bilal; Moussaoui, Abdelouahab

    2018-04-01

    Community structure is very important to understand a network which represents a context. Many community detection methods have been proposed like hierarchical methods. In our study, we propose a new hierarchical method for community detection in networks based on genetic algorithm. In this method we use genetic algorithm to split a network into two networks which maximize the modularity. Each new network represents a cluster (community). Then we repeat the splitting process until we get one node at each cluster. We use the modularity function to measure the strength of the community structure found by our method, which gives us an objective metric for choosing the number of communities into which a network should be divided. We demonstrate that our method are highly effective at discovering community structure in both computer-generated and real-world network data.

  15. Theoretical developments for interpreting kernel spectral clustering from alternative viewpoints

    Directory of Open Access Journals (Sweden)

    Diego Peluffo-Ordóñez

    2017-08-01

    Full Text Available To perform an exploration process over complex structured data within unsupervised settings, the so-called kernel spectral clustering (KSC is one of the most recommended and appealing approaches, given its versatility and elegant formulation. In this work, we explore the relationship between (KSC and other well-known approaches, namely normalized cut clustering and kernel k-means. To do so, we first deduce a generic KSC model from a primal-dual formulation based on least-squares support-vector machines (LS-SVM. For experiments, KSC as well as other consider methods are assessed on image segmentation tasks to prove their usability.

  16. Energy Aware Cluster Based Routing Scheme For Wireless Sensor Network

    Directory of Open Access Journals (Sweden)

    Roy Sohini

    2015-09-01

    Full Text Available Wireless Sensor Network (WSN has emerged as an important supplement to the modern wireless communication systems due to its wide range of applications. The recent researches are facing the various challenges of the sensor network more gracefully. However, energy efficiency has still remained a matter of concern for the researches. Meeting the countless security needs, timely data delivery and taking a quick action, efficient route selection and multi-path routing etc. can only be achieved at the cost of energy. Hierarchical routing is more useful in this regard. The proposed algorithm Energy Aware Cluster Based Routing Scheme (EACBRS aims at conserving energy with the help of hierarchical routing by calculating the optimum number of cluster heads for the network, selecting energy-efficient route to the sink and by offering congestion control. Simulation results prove that EACBRS performs better than existing hierarchical routing algorithms like Distributed Energy-Efficient Clustering (DEEC algorithm for heterogeneous wireless sensor networks and Energy Efficient Heterogeneous Clustered scheme for Wireless Sensor Network (EEHC.

  17. Semi-supervised consensus clustering for gene expression data analysis

    OpenAIRE

    Wang, Yunli; Pan, Youlian

    2014-01-01

    Background Simple clustering methods such as hierarchical clustering and k-means are widely used for gene expression data analysis; but they are unable to deal with noise and high dimensionality associated with the microarray gene expression data. Consensus clustering appears to improve the robustness and quality of clustering results. Incorporating prior knowledge in clustering process (semi-supervised clustering) has been shown to improve the consistency between the data partitioning and do...

  18. Implementasi KD-Tree K-Means Clustering untuk Klasterisasi Dokumen

    Directory of Open Access Journals (Sweden)

    Eric Budiman Gosno

    2013-09-01

    Full Text Available Klasterisasi dokumen adalah suatu proses pengelompokan dokumen secara otomatis dan unsupervised. Klasterisasi dokumen merupakan permasalahan yang sering ditemui dalam berbagai bidang seperti text mining dan sistem temu kembali informasi. Metode klasterisasi dokumen yang memiliki akurasi dan efisiensi waktu yang tinggi sangat diperlukan untuk meningkatkan hasil pada mesin pencari web,  dan untuk proses filtering. Salah satu metode klasterisasi yang telah dikenal dan diaplikasikan dalam klasterisasi dokumen adalah K-Means Clustering. Tetapi K-Means Clustering sensitif terhadap pemilihan posisi awal dari titik tengah klaster sehingga pemilihan posisi awal dari titik tengah klaster yang buruk akan mengakibatkan K-Means Clustering terjebak dalam local optimum. KD-Tree K-Means Clustering merupakan perbaikan dari K-Means Clustering. KD-Tree K-Means Clustering menggunakan struktur data K-Dimensional Tree dan nilai kerapatan pada proses inisialisasi titik tengah klaster. Pada makalah ini diimplementasikan algoritma KD-Tree K-Means Clustering untuk permasalahan klasterisasi dokumen. Performa klasterisasi dokumen yang dihasilkan oleh metode KD-Tree K-Means Clustering pada data set 20 newsgroup memiliki nilai distorsi 3×105 lebih rendah dibandingkan dengan nilai rerata distorsi K-Means Clustering dan nilai NIG 0,09 lebih baik dibandingkan dengan nilai NIG K-Means Clustering.

  19. 3D NEAREST NEIGHBOUR SEARCH USING A CLUSTERED HIERARCHICAL TREE STRUCTURE

    Directory of Open Access Journals (Sweden)

    A. Suhaibah

    2016-06-01

    Full Text Available Locating and analysing the location of new stores or outlets is one of the common issues facing retailers and franchisers. This is due to assure that new opening stores are at their strategic location to attract the highest possible number of customers. Spatial information is used to manage, maintain and analyse these store locations. However, since the business of franchising and chain stores in urban areas runs within high rise multi-level buildings, a three-dimensional (3D method is prominently required in order to locate and identify the surrounding information such as at which level of the franchise unit will be located or is the franchise unit located is at the best level for visibility purposes. One of the common used analyses used for retrieving the surrounding information is Nearest Neighbour (NN analysis. It uses a point location and identifies the surrounding neighbours. However, with the immense number of urban datasets, the retrieval and analysis of nearest neighbour information and their efficiency will become more complex and crucial. In this paper, we present a technique to retrieve nearest neighbour information in 3D space using a clustered hierarchical tree structure. Based on our findings, the proposed approach substantially showed an improvement of response time analysis compared to existing approaches of spatial access methods in databases. The query performance was tested using a dataset consisting of 500,000 point locations building and franchising unit. The results are presented in this paper. Another advantage of this structure is that it also offers a minimal overlap and coverage among nodes which can reduce repetitive data entry.

  20. Hierarchical Cluster-based Partial Least Squares Regression (HC-PLSR is an efficient tool for metamodelling of nonlinear dynamic models

    Directory of Open Access Journals (Sweden)

    Omholt Stig W

    2011-06-01

    Full Text Available Abstract Background Deterministic dynamic models of complex biological systems contain a large number of parameters and state variables, related through nonlinear differential equations with various types of feedback. A metamodel of such a dynamic model is a statistical approximation model that maps variation in parameters and initial conditions (inputs to variation in features of the trajectories of the state variables (outputs throughout the entire biologically relevant input space. A sufficiently accurate mapping can be exploited both instrumentally and epistemically. Multivariate regression methodology is a commonly used approach for emulating dynamic models. However, when the input-output relations are highly nonlinear or non-monotone, a standard linear regression approach is prone to give suboptimal results. We therefore hypothesised that a more accurate mapping can be obtained by locally linear or locally polynomial regression. We present here a new method for local regression modelling, Hierarchical Cluster-based PLS regression (HC-PLSR, where fuzzy C-means clustering is used to separate the data set into parts according to the structure of the response surface. We compare the metamodelling performance of HC-PLSR with polynomial partial least squares regression (PLSR and ordinary least squares (OLS regression on various systems: six different gene regulatory network models with various types of feedback, a deterministic mathematical model of the mammalian circadian clock and a model of the mouse ventricular myocyte function. Results Our results indicate that multivariate regression is well suited for emulating dynamic models in systems biology. The hierarchical approach turned out to be superior to both polynomial PLSR and OLS regression in all three test cases. The advantage, in terms of explained variance and prediction accuracy, was largest in systems with highly nonlinear functional relationships and in systems with positive feedback

  1. Hierarchical cluster-based partial least squares regression (HC-PLSR) is an efficient tool for metamodelling of nonlinear dynamic models.

    Science.gov (United States)

    Tøndel, Kristin; Indahl, Ulf G; Gjuvsland, Arne B; Vik, Jon Olav; Hunter, Peter; Omholt, Stig W; Martens, Harald

    2011-06-01

    Deterministic dynamic models of complex biological systems contain a large number of parameters and state variables, related through nonlinear differential equations with various types of feedback. A metamodel of such a dynamic model is a statistical approximation model that maps variation in parameters and initial conditions (inputs) to variation in features of the trajectories of the state variables (outputs) throughout the entire biologically relevant input space. A sufficiently accurate mapping can be exploited both instrumentally and epistemically. Multivariate regression methodology is a commonly used approach for emulating dynamic models. However, when the input-output relations are highly nonlinear or non-monotone, a standard linear regression approach is prone to give suboptimal results. We therefore hypothesised that a more accurate mapping can be obtained by locally linear or locally polynomial regression. We present here a new method for local regression modelling, Hierarchical Cluster-based PLS regression (HC-PLSR), where fuzzy C-means clustering is used to separate the data set into parts according to the structure of the response surface. We compare the metamodelling performance of HC-PLSR with polynomial partial least squares regression (PLSR) and ordinary least squares (OLS) regression on various systems: six different gene regulatory network models with various types of feedback, a deterministic mathematical model of the mammalian circadian clock and a model of the mouse ventricular myocyte function. Our results indicate that multivariate regression is well suited for emulating dynamic models in systems biology. The hierarchical approach turned out to be superior to both polynomial PLSR and OLS regression in all three test cases. The advantage, in terms of explained variance and prediction accuracy, was largest in systems with highly nonlinear functional relationships and in systems with positive feedback loops. HC-PLSR is a promising approach for

  2. Coherence-based Time Series Clustering for Brain Connectivity Visualization

    KAUST Repository

    Euan, Carolina

    2017-11-19

    We develop the hierarchical cluster coherence (HCC) method for brain signals, a procedure for characterizing connectivity in a network by clustering nodes or groups of channels that display high level of coordination as measured by

  3. Coherence-based Time Series Clustering for Brain Connectivity Visualization

    KAUST Repository

    Euan, Carolina; Sun, Ying; Ombao, Hernando

    2017-01-01

    We develop the hierarchical cluster coherence (HCC) method for brain signals, a procedure for characterizing connectivity in a network by clustering nodes or groups of channels that display high level of coordination as measured by

  4. Programming Hierarchical Self-Assembly of Patchy Particles into Colloidal Crystals via Colloidal Molecules.

    Science.gov (United States)

    Morphew, Daniel; Shaw, James; Avins, Christopher; Chakrabarti, Dwaipayan

    2018-03-27

    Colloidal self-assembly is a promising bottom-up route to a wide variety of three-dimensional structures, from clusters to crystals. Programming hierarchical self-assembly of colloidal building blocks, which can give rise to structures ordered at multiple levels to rival biological complexity, poses a multiscale design problem. Here we explore a generic design principle that exploits a hierarchy of interaction strengths and employ this design principle in computer simulations to demonstrate the hierarchical self-assembly of triblock patchy colloidal particles into two distinct colloidal crystals. We obtain cubic diamond and body-centered cubic crystals via distinct clusters of uniform size and shape, namely, tetrahedra and octahedra, respectively. Such a conceptual design framework has the potential to reliably encode hierarchical self-assembly of colloidal particles into a high level of sophistication. Moreover, the design framework underpins a bottom-up route to cubic diamond colloidal crystals, which have remained elusive despite being much sought after for their attractive photonic applications.

  5. Semi-supervised Probabilistic Distance Clustering and the Uncertainty of Classification

    Science.gov (United States)

    Iyigun, Cem; Ben-Israel, Adi

    Semi-supervised clustering is an attempt to reconcile clustering (unsupervised learning) and classification (supervised learning, using prior information on the data). These two modes of data analysis are combined in a parameterized model, the parameter θ ∈ [0, 1] is the weight attributed to the prior information, θ = 0 corresponding to clustering, and θ = 1 to classification. The results (cluster centers, classification rule) depend on the parameter θ, an insensitivity to θ indicates that the prior information is in agreement with the intrinsic cluster structure, and is otherwise redundant. This explains why some data sets (such as the Wisconsin breast cancer data, Merz and Murphy, UCI repository of machine learning databases, University of California, Irvine, CA) give good results for all reasonable classification methods. The uncertainty of classification is represented here by the geometric mean of the membership probabilities, shown to be an entropic distance related to the Kullback-Leibler divergence.

  6. High-speed detection of emergent market clustering via an unsupervised parallel genetic algorithm

    Directory of Open Access Journals (Sweden)

    Dieter Hendricks

    2016-02-01

    Full Text Available We implement a master-slave parallel genetic algorithm with a bespoke log-likelihood fitness function to identify emergent clusters within price evolutions. We use graphics processing units (GPUs to implement a parallel genetic algorithm and visualise the results using disjoint minimal spanning trees. We demonstrate that our GPU parallel genetic algorithm, implemented on a commercially available general purpose GPU, is able to recover stock clusters in sub-second speed, based on a subset of stocks in the South African market. This approach represents a pragmatic choice for low-cost, scalable parallel computing and is significantly faster than a prototype serial implementation in an optimised C-based fourth-generation programming language, although the results are not directly comparable because of compiler differences. Combined with fast online intraday correlation matrix estimation from high frequency data for cluster identification, the proposed implementation offers cost-effective, near-real-time risk assessment for financial practitioners.

  7. Hierarchical Image Segmentation of Remotely Sensed Data using Massively Parallel GNU-LINUX Software

    Science.gov (United States)

    Tilton, James C.

    2003-01-01

    A hierarchical set of image segmentations is a set of several image segmentations of the same image at different levels of detail in which the segmentations at coarser levels of detail can be produced from simple merges of regions at finer levels of detail. In [1], Tilton, et a1 describes an approach for producing hierarchical segmentations (called HSEG) and gave a progress report on exploiting these hierarchical segmentations for image information mining. The HSEG algorithm is a hybrid of region growing and constrained spectral clustering that produces a hierarchical set of image segmentations based on detected convergence points. In the main, HSEG employs the hierarchical stepwise optimization (HSWO) approach to region growing, which was described as early as 1989 by Beaulieu and Goldberg. The HSWO approach seeks to produce segmentations that are more optimized than those produced by more classic approaches to region growing (e.g. Horowitz and T. Pavlidis, [3]). In addition, HSEG optionally interjects between HSWO region growing iterations, merges between spatially non-adjacent regions (i.e., spectrally based merging or clustering) constrained by a threshold derived from the previous HSWO region growing iteration. While the addition of constrained spectral clustering improves the utility of the segmentation results, especially for larger images, it also significantly increases HSEG s computational requirements. To counteract this, a computationally efficient recursive, divide-and-conquer, implementation of HSEG (RHSEG) was devised, which includes special code to avoid processing artifacts caused by RHSEG s recursive subdivision of the image data. The recursive nature of RHSEG makes for a straightforward parallel implementation. This paper describes the HSEG algorithm, its recursive formulation (referred to as RHSEG), and the implementation of RHSEG using massively parallel GNU-LINUX software. Results with Landsat TM data are included comparing RHSEG with classic

  8. Symmetric nonnegative matrix factorization: algorithms and applications to probabilistic clustering.

    Science.gov (United States)

    He, Zhaoshui; Xie, Shengli; Zdunek, Rafal; Zhou, Guoxu; Cichocki, Andrzej

    2011-12-01

    Nonnegative matrix factorization (NMF) is an unsupervised learning method useful in various applications including image processing and semantic analysis of documents. This paper focuses on symmetric NMF (SNMF), which is a special case of NMF decomposition. Three parallel multiplicative update algorithms using level 3 basic linear algebra subprograms directly are developed for this problem. First, by minimizing the Euclidean distance, a multiplicative update algorithm is proposed, and its convergence under mild conditions is proved. Based on it, we further propose another two fast parallel methods: α-SNMF and β -SNMF algorithms. All of them are easy to implement. These algorithms are applied to probabilistic clustering. We demonstrate their effectiveness for facial image clustering, document categorization, and pattern clustering in gene expression.

  9. Usage of self-organizing neural networks in evaluation of consumer behaviour

    Directory of Open Access Journals (Sweden)

    Jana Weinlichová

    2010-01-01

    Full Text Available This article deals with evaluation of consumer data by Artificial Intelligence methods. In methodical part there are described learning algorithms for Kohonen maps on the principle of supervised learning, unsupervised learning and semi-supervised learning. The principles of supervised learning and unsupervised learning are compared. On base of binding conditions of these principles there is pointed out an advantage of semi-supervised learning. Three algorithms are described for the semi-supervised learning: label propagation, self-training and co-training. Especially usage of co-training in Kohonen map learning seems to be promising point of other research. In concrete application of Kohonen neural network on consumer’s expense the unsupervised learning method has been chosen – the self-organization. So the features of data are evaluated by clustering method called Kohonen maps. These input data represents consumer expenses of households in countries of European union and are characterised by 12-dimension vector according to commodity classification. The data are evaluated in several years, so we can see their distribution, similarity or dissimilarity and also their evolution. In the article we discus other usage of this method for this type of data and also comparison of our results with results reached by hierarchical cluster analysis.

  10. Unsupervised segmentation of lung fields in chest radiographs using multiresolution fractal feature vector and deformable models.

    Science.gov (United States)

    Lee, Wen-Li; Chang, Koyin; Hsieh, Kai-Sheng

    2016-09-01

    Segmenting lung fields in a chest radiograph is essential for automatically analyzing an image. We present an unsupervised method based on multiresolution fractal feature vector. The feature vector characterizes the lung field region effectively. A fuzzy c-means clustering algorithm is then applied to obtain a satisfactory initial contour. The final contour is obtained by deformable models. The results show the feasibility and high performance of the proposed method. Furthermore, based on the segmentation of lung fields, the cardiothoracic ratio (CTR) can be measured. The CTR is a simple index for evaluating cardiac hypertrophy. After identifying a suspicious symptom based on the estimated CTR, a physician can suggest that the patient undergoes additional extensive tests before a treatment plan is finalized.

  11. Automated Glioblastoma Segmentation Based on a Multiparametric Structured Unsupervised Classification

    Science.gov (United States)

    Juan-Albarracín, Javier; Fuster-Garcia, Elies; Manjón, José V.; Robles, Montserrat; Aparici, F.; Martí-Bonmatí, L.; García-Gómez, Juan M.

    2015-01-01

    Automatic brain tumour segmentation has become a key component for the future of brain tumour treatment. Currently, most of brain tumour segmentation approaches arise from the supervised learning standpoint, which requires a labelled training dataset from which to infer the models of the classes. The performance of these models is directly determined by the size and quality of the training corpus, whose retrieval becomes a tedious and time-consuming task. On the other hand, unsupervised approaches avoid these limitations but often do not reach comparable results than the supervised methods. In this sense, we propose an automated unsupervised method for brain tumour segmentation based on anatomical Magnetic Resonance (MR) images. Four unsupervised classification algorithms, grouped by their structured or non-structured condition, were evaluated within our pipeline. Considering the non-structured algorithms, we evaluated K-means, Fuzzy K-means and Gaussian Mixture Model (GMM), whereas as structured classification algorithms we evaluated Gaussian Hidden Markov Random Field (GHMRF). An automated postprocess based on a statistical approach supported by tissue probability maps is proposed to automatically identify the tumour classes after the segmentations. We evaluated our brain tumour segmentation method with the public BRAin Tumor Segmentation (BRATS) 2013 Test and Leaderboard datasets. Our approach based on the GMM model improves the results obtained by most of the supervised methods evaluated with the Leaderboard set and reaches the second position in the ranking. Our variant based on the GHMRF achieves the first position in the Test ranking of the unsupervised approaches and the seventh position in the general Test ranking, which confirms the method as a viable alternative for brain tumour segmentation. PMID:25978453

  12. Genomic signal processing for DNA sequence clustering.

    Science.gov (United States)

    Mendizabal-Ruiz, Gerardo; Román-Godínez, Israel; Torres-Ramos, Sulema; Salido-Ruiz, Ricardo A; Vélez-Pérez, Hugo; Morales, J Alejandro

    2018-01-01

    Genomic signal processing (GSP) methods which convert DNA data to numerical values have recently been proposed, which would offer the opportunity of employing existing digital signal processing methods for genomic data. One of the most used methods for exploring data is cluster analysis which refers to the unsupervised classification of patterns in data. In this paper, we propose a novel approach for performing cluster analysis of DNA sequences that is based on the use of GSP methods and the K-means algorithm. We also propose a visualization method that facilitates the easy inspection and analysis of the results and possible hidden behaviors. Our results support the feasibility of employing the proposed method to find and easily visualize interesting features of sets of DNA data.

  13. Dynamic Hierarchical Energy-Efficient Method Based on Combinatorial Optimization for Wireless Sensor Networks.

    Science.gov (United States)

    Chang, Yuchao; Tang, Hongying; Cheng, Yongbo; Zhao, Qin; Yuan, Baoqing Li andXiaobing

    2017-07-19

    Routing protocols based on topology control are significantly important for improving network longevity in wireless sensor networks (WSNs). Traditionally, some WSN routing protocols distribute uneven network traffic load to sensor nodes, which is not optimal for improving network longevity. Differently to conventional WSN routing protocols, we propose a dynamic hierarchical protocol based on combinatorial optimization (DHCO) to balance energy consumption of sensor nodes and to improve WSN longevity. For each sensor node, the DHCO algorithm obtains the optimal route by establishing a feasible routing set instead of selecting the cluster head or the next hop node. The process of obtaining the optimal route can be formulated as a combinatorial optimization problem. Specifically, the DHCO algorithm is carried out by the following procedures. It employs a hierarchy-based connection mechanism to construct a hierarchical network structure in which each sensor node is assigned to a special hierarchical subset; it utilizes the combinatorial optimization theory to establish the feasible routing set for each sensor node, and takes advantage of the maximum-minimum criterion to obtain their optimal routes to the base station. Various results of simulation experiments show effectiveness and superiority of the DHCO algorithm in comparison with state-of-the-art WSN routing algorithms, including low-energy adaptive clustering hierarchy (LEACH), hybrid energy-efficient distributed clustering (HEED), genetic protocol-based self-organizing network clustering (GASONeC), and double cost function-based routing (DCFR) algorithms.

  14. Cluster Validity Classification Approaches Based on Geometric Probability and Application in the Classification of Remotely Sensed Images

    Directory of Open Access Journals (Sweden)

    LI Jian-Wei

    2014-08-01

    Full Text Available On the basis of the cluster validity function based on geometric probability in literature [1, 2], propose a cluster analysis method based on geometric probability to process large amount of data in rectangular area. The basic idea is top-down stepwise refinement, firstly categories then subcategories. On all clustering levels, use the cluster validity function based on geometric probability firstly, determine clusters and the gathering direction, then determine the center of clustering and the border of clusters. Through TM remote sensing image classification examples, compare with the supervision and unsupervised classification in ERDAS and the cluster analysis method based on geometric probability in two-dimensional square which is proposed in literature 2. Results show that the proposed method can significantly improve the classification accuracy.

  15. Clustering analysis of malware behavior using Self Organizing Map

    DEFF Research Database (Denmark)

    Pirscoveanu, Radu-Stefan; Stevanovic, Matija; Pedersen, Jens Myrup

    2016-01-01

    For the time being, malware behavioral classification is performed by means of Anti-Virus (AV) generated labels. The paper investigates the inconsistencies associated with current practices by evaluating the identified differences between current vendors. In this paper we rely on Self Organizing...... Map, an unsupervised machine learning algorithm, for generating clusters that capture the similarities between malware behavior. A data set of approximately 270,000 samples was used to generate the behavioral profile of malicious types in order to compare the outcome of the proposed clustering...... approach with the labels collected from 57 Antivirus vendors using VirusTotal. Upon evaluating the results, the paper concludes on shortcomings of relying on AV vendors for labeling malware samples. In order to solve the problem, a cluster-based classification is proposed, which should provide more...

  16. Chimera states in networks of logistic maps with hierarchical connectivities

    Science.gov (United States)

    zur Bonsen, Alexander; Omelchenko, Iryna; Zakharova, Anna; Schöll, Eckehard

    2018-04-01

    Chimera states are complex spatiotemporal patterns consisting of coexisting domains of coherence and incoherence. We study networks of nonlocally coupled logistic maps and analyze systematically how the dilution of the network links influences the appearance of chimera patterns. The network connectivities are constructed using an iterative Cantor algorithm to generate fractal (hierarchical) connectivities. Increasing the hierarchical level of iteration, we compare the resulting spatiotemporal patterns. We demonstrate that a high clustering coefficient and symmetry of the base pattern promotes chimera states, and asymmetric connectivities result in complex nested chimera patterns.

  17. Ultrathin mesoporous Co{sub 3}O{sub 4} nanosheets-constructed hierarchical clusters as high rate capability and long life anode materials for lithium-ion batteries

    Energy Technology Data Exchange (ETDEWEB)

    Wu, Shengming [Key Laboratory of Functional Inorganic Materials Chemistry, Ministry of Education, School of Chemistry, Chemical Engineering and Materials, Heilongjiang University, Heilongjiang, Harbin 150080 (China); Xia, Tian, E-mail: xiatian@hlju.edu.cn [Key Laboratory of Functional Inorganic Materials Chemistry, Ministry of Education, School of Chemistry, Chemical Engineering and Materials, Heilongjiang University, Heilongjiang, Harbin 150080 (China); Wang, Jingping [Key Laboratory of Superlight Material and Surface Technology, Ministry of Education, College of Materials Science and Chemical Engineering, Harbin Engineering University, Heilongjiang, Harbin 150001 (China); Lu, Feifei [Key Laboratory of Functional Inorganic Materials Chemistry, Ministry of Education, School of Chemistry, Chemical Engineering and Materials, Heilongjiang University, Heilongjiang, Harbin 150080 (China); Xu, Chunbo [Key Laboratory of Superlight Material and Surface Technology, Ministry of Education, College of Materials Science and Chemical Engineering, Harbin Engineering University, Heilongjiang, Harbin 150001 (China); Zhang, Xianfa; Huo, Lihua [Key Laboratory of Functional Inorganic Materials Chemistry, Ministry of Education, School of Chemistry, Chemical Engineering and Materials, Heilongjiang University, Heilongjiang, Harbin 150080 (China); Zhao, Hui, E-mail: zhaohui98@yahoo.com [Key Laboratory of Functional Inorganic Materials Chemistry, Ministry of Education, School of Chemistry, Chemical Engineering and Materials, Heilongjiang University, Heilongjiang, Harbin 150080 (China)

    2017-06-01

    Graphical abstract: Ultrathin mesoporous Co{sub 3}O{sub 4} nanosheets-constructed hierarchical clusters (UMCN-HCs) have been successfully synthesized via a facile hydrothermal method followed by a subsequent thermolysis treatment. When tested as anode materials for LIBs, UMCN-HCs achieve high reversible capacity, good long cycling life, and rate capability. - Highlights: • UMCN-HCs show high capacity, excellent stability, and good rate capability. • UMCN-HCs retain a capacity of 1067 mAh g{sup −1} after 100 cycles at 100 mA g{sup −1}. • UMCN-HCs deliver a capacity of 507 mAh g{sup −1} after 500 cycles at 2 A g{sup −1}. - Abstract: Herein, Ultrathin mesoporous Co{sub 3}O{sub 4} nanosheets-constructed hierarchical clusters (UMCN-HCs) have been successfully synthesized via a facile hydrothermal method followed by a subsequent thermolysis treatment at 600 °C in air. The products consist of cluster-like Co{sub 3}O{sub 4} microarchitectures, which are assembled by numerous ultrathin mesoporous Co{sub 3}O{sub 4} nanosheets. When tested as anode materials for lithium-ion batteries, UMCN-HCs deliver a high reversible capacity of 1067 mAh g{sup −1} at a current density of 100 mA g{sup −1} after 100 cycles. Even at 2 A g{sup −1}, a stable capacity as high as 507 mAh g{sup −1} can be achieved after 500 cycles. The high reversible capacity, excellent cycling stability, and good rate capability of UMCN-HCs may be attributed to their mesoporous sheet-like nanostructure. The sheet-layered structure of UMCN-HCs may buffer the volume change during the lithiation-delithiation process, and the mesoporous characteristic make lithium-ion transfer more easily at the interface between the active electrode and the electrolyte.

  18. Energy Aware Clustering Algorithms for Wireless Sensor Networks

    Science.gov (United States)

    Rakhshan, Noushin; Rafsanjani, Marjan Kuchaki; Liu, Chenglian

    2011-09-01

    The sensor nodes deployed in wireless sensor networks (WSNs) are extremely power constrained, so maximizing the lifetime of the entire networks is mainly considered in the design. In wireless sensor networks, hierarchical network structures have the advantage of providing scalable and energy efficient solutions. In this paper, we investigate different clustering algorithms for WSNs and also compare these clustering algorithms based on metrics such as clustering distribution, cluster's load balancing, Cluster Head's (CH) selection strategy, CH's role rotation, node mobility, clusters overlapping, intra-cluster communications, reliability, security and location awareness.

  19. MAP-Based Underdetermined Blind Source Separation of Convolutive Mixtures by Hierarchical Clustering and -Norm Minimization

    Directory of Open Access Journals (Sweden)

    Kellermann Walter

    2007-01-01

    Full Text Available We address the problem of underdetermined BSS. While most previous approaches are designed for instantaneous mixtures, we propose a time-frequency-domain algorithm for convolutive mixtures. We adopt a two-step method based on a general maximum a posteriori (MAP approach. In the first step, we estimate the mixing matrix based on hierarchical clustering, assuming that the source signals are sufficiently sparse. The algorithm works directly on the complex-valued data in the time-frequency domain and shows better convergence than algorithms based on self-organizing maps. The assumption of Laplacian priors for the source signals in the second step leads to an algorithm for estimating the source signals. It involves the -norm minimization of complex numbers because of the use of the time-frequency-domain approach. We compare a combinatorial approach initially designed for real numbers with a second-order cone programming (SOCP approach designed for complex numbers. We found that although the former approach is not theoretically justified for complex numbers, its results are comparable to, or even better than, the SOCP solution. The advantage is a lower computational cost for problems with low input/output dimensions.

  20. Hierarchical structure in the distribution of galaxies

    International Nuclear Information System (INIS)

    Schulman, L.S.; Seiden, P.E.; Technion - Israel Institute of Technology, Haifa; IBM Thomas J. Watson Research Center, Yorktown Heights, NY)

    1986-01-01

    The distribution of galaxies has a hierarchical structure with power-law correlations. This is usually thought to arise from gravity alone acting on an originally uniform distributioon. If, however, the original process of galaxy formation occurs through the stimulated birth of one galaxy due to a nearby recently formed galaxy, and if this process occurs near its percolation threshold, then a hierarchical structure with power-law correlations arises at the time of galaxy formation. If subsequent gravitational evolution within an expanding cosmology is such as to retain power-law correlations, the initial r exp -1 dropoff can steepen to the observed r exp -1.8. The distribution of galaxies obtained by this process produces clustering and voids, as observed. 23 references

  1. A CLUSTER IN THE MAKING: ALMA REVEALS THE INITIAL CONDITIONS FOR HIGH-MASS CLUSTER FORMATION

    International Nuclear Information System (INIS)

    Rathborne, J. M.; Contreras, Y.; Longmore, S. N.; Bastian, N.; Jackson, J. M.; Alves, J. F.; Bally, J.; Foster, J. B.; Garay, G.; Kruijssen, J. M. D.; Testi, L.; Walsh, A. J.

    2015-01-01

    G0.253+0.016 is a molecular clump that appears to be on the verge of forming a high-mass cluster: its extremely low dust temperature, high mass, and high density, combined with its lack of prevalent star formation, make it an excellent candidate for an Arches-like cluster in a very early stage of formation. Here we present new Atacama Large Millimeter/Sub-millimeter Array observations of its small-scale (∼0.07 pc) 3 mm dust continuum and molecular line emission from 17 different species that probe a range of distinct physical and chemical conditions. The data reveal a complex network of emission features with a complicated velocity structure: there is emission on all spatial scales, the morphology of which ranges from small, compact regions to extended, filamentary structures that are seen in both emission and absorption. The dust column density is well traced by molecules with higher excitation energies and critical densities, consistent with a clump that has a denser interior. A statistical analysis supports the idea that turbulence shapes the observed gas structure within G0.253+0.016. We find a clear break in the turbulent power spectrum derived from the optically thin dust continuum emission at a spatial scale of ∼0.1 pc, which may correspond to the spatial scale at which gravity has overcome the thermal pressure. We suggest that G0.253+0.016 is on the verge of forming a cluster from hierarchical, filamentary structures that arise from a highly turbulent medium. Although the stellar distribution within high-mass Arches-like clusters is compact, centrally condensed, and smooth, the observed gas distribution within G0.253+0.016 is extended, with no high-mass central concentration, and has a complex, hierarchical structure. If this clump gives rise to a high-mass cluster and its stars are formed from this initially hierarchical gas structure, then the resulting cluster must evolve into a centrally condensed structure via a dynamical process

  2. Unsupervised learning of facial emotion decoding skills

    Directory of Open Access Journals (Sweden)

    Jan Oliver Huelle

    2014-02-01

    Full Text Available Research on the mechanisms underlying human facial emotion recognition has long focussed on genetically determined neural algorithms and often neglected the question of how these algorithms might be tuned by social learning. Here we show that facial emotion decoding skills can be significantly and sustainably improved by practise without an external teaching signal. Participants saw video clips of dynamic facial expressions of five different women and were asked to decide which of four possible emotions (anger, disgust, fear and sadness was shown in each clip. Although no external information about the correctness of the participant’s response or the sender’s true affective state was provided, participants showed a significant increase of facial emotion recognition accuracy both within and across two training sessions two days to several weeks apart. We discuss several similarities and differences between the unsupervised improvement of facial decoding skills observed in the current study, unsupervised perceptual learning of simple stimuli described in previous studies and practise effects often observed in cognitive tasks.

  3. Unsupervised Learning —A Novel Clustering Method for Rolling Bearing Faults Identification

    Science.gov (United States)

    Kai, Li; Bo, Luo; Tao, Ma; Xuefeng, Yang; Guangming, Wang

    2017-12-01

    To promptly process the massive fault data and automatically provide accurate diagnosis results, numerous studies have been conducted on intelligent fault diagnosis of rolling bearing. Among these studies, such as artificial neural networks, support vector machines, decision trees and other supervised learning methods are used commonly. These methods can detect the failure of rolling bearing effectively, but to achieve better detection results, it often requires a lot of training samples. Based on above, a novel clustering method is proposed in this paper. This novel method is able to find the correct number of clusters automatically the effectiveness of the proposed method is validated using datasets from rolling element bearings. The diagnosis results show that the proposed method can accurately detect the fault types of small samples. Meanwhile, the diagnosis results are also relative high accuracy even for massive samples.

  4. Cluster Physics with Merging Galaxy Clusters

    Directory of Open Access Journals (Sweden)

    Sandor M. Molnar

    2016-02-01

    Full Text Available Collisions between galaxy clusters provide a unique opportunity to study matter in a parameter space which cannot be explored in our laboratories on Earth. In the standard LCDM model, where the total density is dominated by the cosmological constant ($Lambda$ and the matter density by cold dark matter (CDM, structure formation is hierarchical, and clusters grow mostly by merging.Mergers of two massive clusters are the most energetic events in the universe after the Big Bang,hence they provide a unique laboratory to study cluster physics.The two main mass components in clusters behave differently during collisions:the dark matter is nearly collisionless, responding only to gravity, while the gas is subject to pressure forces and dissipation, and shocks and turbulenceare developed during collisions. In the present contribution we review the different methods used to derive the physical properties of merging clusters. Different physical processes leave their signatures on different wavelengths, thusour review is based on a multifrequency analysis. In principle, the best way to analyze multifrequency observations of merging clustersis to model them using N-body/HYDRO numerical simulations. We discuss the results of such detailed analyses.New high spatial and spectral resolution ground and space based telescopeswill come online in the near future. Motivated by these new opportunities,we briefly discuss methods which will be feasible in the near future in studying merging clusters.

  5. A Genetic Algorithm That Exchanges Neighboring Centers for Fuzzy c-Means Clustering

    Science.gov (United States)

    Chahine, Firas Safwan

    2012-01-01

    Clustering algorithms are widely used in pattern recognition and data mining applications. Due to their computational efficiency, partitional clustering algorithms are better suited for applications with large datasets than hierarchical clustering algorithms. K-means is among the most popular partitional clustering algorithm, but has a major…

  6. A competition in unsupervised color image segmentation

    Czech Academy of Sciences Publication Activity Database

    Haindl, Michal; Mikeš, Stanislav

    2016-01-01

    Roč. 57, č. 9 (2016), s. 136-151 ISSN 0031-3203 R&D Projects: GA ČR(CZ) GA14-10911S Institutional support: RVO:67985556 Keywords : Unsupervised image segmentation * Segmentation contest * Texture analysis Subject RIV: BD - Theory of Information Impact factor: 4.582, year: 2016 http://library.utia.cas.cz/separaty/2016/RO/haindl-0459179.pdf

  7. A Novel Unsupervised Segmentation Quality Evaluation Method for Remote Sensing Images.

    Science.gov (United States)

    Gao, Han; Tang, Yunwei; Jing, Linhai; Li, Hui; Ding, Haifeng

    2017-10-24

    The segmentation of a high spatial resolution remote sensing image is a critical step in geographic object-based image analysis (GEOBIA). Evaluating the performance of segmentation without ground truth data, i.e., unsupervised evaluation, is important for the comparison of segmentation algorithms and the automatic selection of optimal parameters. This unsupervised strategy currently faces several challenges in practice, such as difficulties in designing effective indicators and limitations of the spectral values in the feature representation. This study proposes a novel unsupervised evaluation method to quantitatively measure the quality of segmentation results to overcome these problems. In this method, multiple spectral and spatial features of images are first extracted simultaneously and then integrated into a feature set to improve the quality of the feature representation of ground objects. The indicators designed for spatial stratified heterogeneity and spatial autocorrelation are included to estimate the properties of the segments in this integrated feature set. These two indicators are then combined into a global assessment metric as the final quality score. The trade-offs of the combined indicators are accounted for using a strategy based on the Mahalanobis distance, which can be exhibited geometrically. The method is tested on two segmentation algorithms and three testing images. The proposed method is compared with two existing unsupervised methods and a supervised method to confirm its capabilities. Through comparison and visual analysis, the results verified the effectiveness of the proposed method and demonstrated the reliability and improvements of this method with respect to other methods.

  8. A Novel Unsupervised Segmentation Quality Evaluation Method for Remote Sensing Images

    Directory of Open Access Journals (Sweden)

    Han Gao

    2017-10-01

    Full Text Available The segmentation of a high spatial resolution remote sensing image is a critical step in geographic object-based image analysis (GEOBIA. Evaluating the performance of segmentation without ground truth data, i.e., unsupervised evaluation, is important for the comparison of segmentation algorithms and the automatic selection of optimal parameters. This unsupervised strategy currently faces several challenges in practice, such as difficulties in designing effective indicators and limitations of the spectral values in the feature representation. This study proposes a novel unsupervised evaluation method to quantitatively measure the quality of segmentation results to overcome these problems. In this method, multiple spectral and spatial features of images are first extracted simultaneously and then integrated into a feature set to improve the quality of the feature representation of ground objects. The indicators designed for spatial stratified heterogeneity and spatial autocorrelation are included to estimate the properties of the segments in this integrated feature set. These two indicators are then combined into a global assessment metric as the final quality score. The trade-offs of the combined indicators are accounted for using a strategy based on the Mahalanobis distance, which can be exhibited geometrically. The method is tested on two segmentation algorithms and three testing images. The proposed method is compared with two existing unsupervised methods and a supervised method to confirm its capabilities. Through comparison and visual analysis, the results verified the effectiveness of the proposed method and demonstrated the reliability and improvements of this method with respect to other methods.

  9. Efficient Record Linkage Algorithms Using Complete Linkage Clustering.

    Science.gov (United States)

    Mamun, Abdullah-Al; Aseltine, Robert; Rajasekaran, Sanguthevar

    2016-01-01

    Data from different agencies share data of the same individuals. Linking these datasets to identify all the records belonging to the same individuals is a crucial and challenging problem, especially given the large volumes of data. A large number of available algorithms for record linkage are prone to either time inefficiency or low-accuracy in finding matches and non-matches among the records. In this paper we propose efficient as well as reliable sequential and parallel algorithms for the record linkage problem employing hierarchical clustering methods. We employ complete linkage hierarchical clustering algorithms to address this problem. In addition to hierarchical clustering, we also use two other techniques: elimination of duplicate records and blocking. Our algorithms use sorting as a sub-routine to identify identical copies of records. We have tested our algorithms on datasets with millions of synthetic records. Experimental results show that our algorithms achieve nearly 100% accuracy. Parallel implementations achieve almost linear speedups. Time complexities of these algorithms do not exceed those of previous best-known algorithms. Our proposed algorithms outperform previous best-known algorithms in terms of accuracy consuming reasonable run times.

  10. Multi person detection and tracking based on hierarchical level-set method

    Science.gov (United States)

    Khraief, Chadia; Benzarti, Faouzi; Amiri, Hamid

    2018-04-01

    In this paper, we propose an efficient unsupervised method for mutli-person tracking based on hierarchical level-set approach. The proposed method uses both edge and region information in order to effectively detect objects. The persons are tracked on each frame of the sequence by minimizing an energy functional that combines color, texture and shape information. These features are enrolled in covariance matrix as region descriptor. The present method is fully automated without the need to manually specify the initial contour of Level-set. It is based on combined person detection and background subtraction methods. The edge-based is employed to maintain a stable evolution, guide the segmentation towards apparent boundaries and inhibit regions fusion. The computational cost of level-set is reduced by using narrow band technique. Many experimental results are performed on challenging video sequences and show the effectiveness of the proposed method.

  11. Improved regional-scale Brazilian cropping systems' mapping based on a semi-automatic object-based clustering approach

    Science.gov (United States)

    Bellón, Beatriz; Bégué, Agnès; Lo Seen, Danny; Lebourgeois, Valentine; Evangelista, Balbino Antônio; Simões, Margareth; Demonte Ferraz, Rodrigo Peçanha

    2018-06-01

    Cropping systems' maps at fine scale over large areas provide key information for further agricultural production and environmental impact assessments, and thus represent a valuable tool for effective land-use planning. There is, therefore, a growing interest in mapping cropping systems in an operational manner over large areas, and remote sensing approaches based on vegetation index time series analysis have proven to be an efficient tool. However, supervised pixel-based approaches are commonly adopted, requiring resource consuming field campaigns to gather training data. In this paper, we present a new object-based unsupervised classification approach tested on an annual MODIS 16-day composite Normalized Difference Vegetation Index time series and a Landsat 8 mosaic of the State of Tocantins, Brazil, for the 2014-2015 growing season. Two variants of the approach are compared: an hyperclustering approach, and a landscape-clustering approach involving a previous stratification of the study area into landscape units on which the clustering is then performed. The main cropping systems of Tocantins, characterized by the crop types and cropping patterns, were efficiently mapped with the landscape-clustering approach. Results show that stratification prior to clustering significantly improves the classification accuracies for underrepresented and sparsely distributed cropping systems. This study illustrates the potential of unsupervised classification for large area cropping systems' mapping and contributes to the development of generic tools for supporting large-scale agricultural monitoring across regions.

  12. Unsupervised learning of a steerable basis for invariant image representations

    Science.gov (United States)

    Bethge, Matthias; Gerwinn, Sebastian; Macke, Jakob H.

    2007-02-01

    There are two aspects to unsupervised learning of invariant representations of images: First, we can reduce the dimensionality of the representation by finding an optimal trade-off between temporal stability and informativeness. We show that the answer to this optimization problem is generally not unique so that there is still considerable freedom in choosing a suitable basis. Which of the many optimal representations should be selected? Here, we focus on this second aspect, and seek to find representations that are invariant under geometrical transformations occuring in sequences of natural images. We utilize ideas of 'steerability' and Lie groups, which have been developed in the context of filter design. In particular, we show how an anti-symmetric version of canonical correlation analysis can be used to learn a full-rank image basis which is steerable with respect to rotations. We provide a geometric interpretation of this algorithm by showing that it finds the two-dimensional eigensubspaces of the average bivector. For data which exhibits a variety of transformations, we develop a bivector clustering algorithm, which we use to learn a basis of generalized quadrature pairs (i.e. 'complex cells') from sequences of natural images.

  13. Non-Hierarchical Clustering as a method to analyse an open-ended ...

    African Journals Online (AJOL)

    We show that the use of non-hierarchical analysis allows us to interpret the reasoning of students solving different mathematical problems using Algebra, and to separate them into different groups, that can be recognised and characterised by common traits in their answers, without any prior knowledge on the part of the ...

  14. Fuzzy clustering-based segmented attenuation correction in whole-body PET

    CERN Document Server

    Zaidi, H; Boudraa, A; Slosman, DO

    2001-01-01

    Segmented-based attenuation correction is now a widely accepted technique to reduce noise contribution of measured attenuation correction. In this paper, we present a new method for segmenting transmission images in positron emission tomography. This reduces the noise on the correction maps while still correcting for differing attenuation coefficients of specific tissues. Based on the Fuzzy C-Means (FCM) algorithm, the method segments the PET transmission images into a given number of clusters to extract specific areas of differing attenuation such as air, the lungs and soft tissue, preceded by a median filtering procedure. The reconstructed transmission image voxels are therefore segmented into populations of uniform attenuation based on the human anatomy. The clustering procedure starts with an over-specified number of clusters followed by a merging process to group clusters with similar properties and remove some undesired substructures using anatomical knowledge. The method is unsupervised, adaptive and a...

  15. Open source clustering software.

    Science.gov (United States)

    de Hoon, M J L; Imoto, S; Nolan, J; Miyano, S

    2004-06-12

    We have implemented k-means clustering, hierarchical clustering and self-organizing maps in a single multipurpose open-source library of C routines, callable from other C and C++ programs. Using this library, we have created an improved version of Michael Eisen's well-known Cluster program for Windows, Mac OS X and Linux/Unix. In addition, we generated a Python and a Perl interface to the C Clustering Library, thereby combining the flexibility of a scripting language with the speed of C. The C Clustering Library and the corresponding Python C extension module Pycluster were released under the Python License, while the Perl module Algorithm::Cluster was released under the Artistic License. The GUI code Cluster 3.0 for Windows, Macintosh and Linux/Unix, as well as the corresponding command-line program, were released under the same license as the original Cluster code. The complete source code is available at http://bonsai.ims.u-tokyo.ac.jp/mdehoon/software/cluster. Alternatively, Algorithm::Cluster can be downloaded from CPAN, while Pycluster is also available as part of the Biopython distribution.

  16. Hierarchical organization in the temporal structure of infant-direct speech and song.

    Science.gov (United States)

    Falk, Simone; Kello, Christopher T

    2017-06-01

    Caregivers alter the temporal structure of their utterances when talking and singing to infants compared with adult communication. The present study tested whether temporal variability in infant-directed registers serves to emphasize the hierarchical temporal structure of speech. Fifteen German-speaking mothers sang a play song and told a story to their 6-months-old infants, or to an adult. Recordings were analyzed using a recently developed method that determines the degree of nested clustering of temporal events in speech. Events were defined as peaks in the amplitude envelope, and clusters of various sizes related to periods of acoustic speech energy at varying timescales. Infant-directed speech and song clearly showed greater event clustering compared with adult-directed registers, at multiple timescales of hundreds of milliseconds to tens of seconds. We discuss the relation of this newly discovered acoustic property to temporal variability in linguistic units and its potential implications for parent-infant communication and infants learning the hierarchical structures of speech and language. Copyright © 2017 Elsevier B.V. All rights reserved.

  17. Using Cluster Analysis to Compartmentalize a Large Managed Wetland Based on Physical, Biological, and Climatic Geospatial Attributes.

    Science.gov (United States)

    Hahus, Ian; Migliaccio, Kati; Douglas-Mankin, Kyle; Klarenberg, Geraldine; Muñoz-Carpena, Rafael

    2018-04-27

    Hierarchical and partitional cluster analyses were used to compartmentalize Water Conservation Area 1, a managed wetland within the Arthur R. Marshall Loxahatchee National Wildlife Refuge in southeast Florida, USA, based on physical, biological, and climatic geospatial attributes. Single, complete, average, and Ward's linkages were tested during the hierarchical cluster analyses, with average linkage providing the best results. In general, the partitional method, partitioning around medoids, found clusters that were more evenly sized and more spatially aggregated than those resulting from the hierarchical analyses. However, hierarchical analysis appeared to be better suited to identify outlier regions that were significantly different from other areas. The clusters identified by geospatial attributes were similar to clusters developed for the interior marsh in a separate study using water quality attributes, suggesting that similar factors have influenced variations in both the set of physical, biological, and climatic attributes selected in this study and water quality parameters. However, geospatial data allowed further subdivision of several interior marsh clusters identified from the water quality data, potentially indicating zones with important differences in function. Identification of these zones can be useful to managers and modelers by informing the distribution of monitoring equipment and personnel as well as delineating regions that may respond similarly to future changes in management or climate.

  18. Unsupervised Video Shot Detection Using Clustering Ensemble with a Color Global Scale-Invariant Feature Transform Descriptor

    Directory of Open Access Journals (Sweden)

    Yuchou Chang

    2008-02-01

    Full Text Available Scale-invariant feature transform (SIFT transforms a grayscale image into scale-invariant coordinates of local features that are invariant to image scale, rotation, and changing viewpoints. Because of its scale-invariant properties, SIFT has been successfully used for object recognition and content-based image retrieval. The biggest drawback of SIFT is that it uses only grayscale information and misses important visual information regarding color. In this paper, we present the development of a novel color feature extraction algorithm that addresses this problem, and we also propose a new clustering strategy using clustering ensembles for video shot detection. Based on Fibonacci lattice-quantization, we develop a novel color global scale-invariant feature transform (CGSIFT for better description of color contents in video frames for video shot detection. CGSIFT first quantizes a color image, representing it with a small number of color indices, and then uses SIFT to extract features from the quantized color index image. We also develop a new space description method using small image regions to represent global color features as the second step of CGSIFT. Clustering ensembles focusing on knowledge reuse are then applied to obtain better clustering results than using single clustering methods for video shot detection. Evaluation of the proposed feature extraction algorithm and the new clustering strategy using clustering ensembles reveals very promising results for video shot detection.

  19. Unsupervised Video Shot Detection Using Clustering Ensemble with a Color Global Scale-Invariant Feature Transform Descriptor

    Directory of Open Access Journals (Sweden)

    Hong Yi

    2008-01-01

    Full Text Available Abstract Scale-invariant feature transform (SIFT transforms a grayscale image into scale-invariant coordinates of local features that are invariant to image scale, rotation, and changing viewpoints. Because of its scale-invariant properties, SIFT has been successfully used for object recognition and content-based image retrieval. The biggest drawback of SIFT is that it uses only grayscale information and misses important visual information regarding color. In this paper, we present the development of a novel color feature extraction algorithm that addresses this problem, and we also propose a new clustering strategy using clustering ensembles for video shot detection. Based on Fibonacci lattice-quantization, we develop a novel color global scale-invariant feature transform (CGSIFT for better description of color contents in video frames for video shot detection. CGSIFT first quantizes a color image, representing it with a small number of color indices, and then uses SIFT to extract features from the quantized color index image. We also develop a new space description method using small image regions to represent global color features as the second step of CGSIFT. Clustering ensembles focusing on knowledge reuse are then applied to obtain better clustering results than using single clustering methods for video shot detection. Evaluation of the proposed feature extraction algorithm and the new clustering strategy using clustering ensembles reveals very promising results for video shot detection.

  20. Spectral identification of plant communities for mapping of semi-natural grasslands

    DEFF Research Database (Denmark)

    Jacobsen, Anne; Nielsen, Allan Aasbjerg; Ejrnæs, Rasmus

    2000-01-01

    identification of plant communities was based on a hierarchical approach relating the test sites to i) management (Ma) and ii) flora (Fl) using spectral consistency and separability as the main criteria. Evaluation of spectral consistency was based on unsupervised clustering of test sites of Ma classes 1 to 7...... as a measure of plant community heterogeneity within management classes. The spectral analysis as well as the maximum likelihood classification indicated that the source of spectral variation within management classes might be related to vegetation composition....

  1. Learning from label proportions in brain-computer interfaces: Online unsupervised learning with guarantees

    Science.gov (United States)

    Verhoeven, Thibault; Schmid, Konstantin; Müller, Klaus-Robert; Tangermann, Michael; Kindermans, Pieter-Jan

    2017-01-01

    Objective Using traditional approaches, a brain-computer interface (BCI) requires the collection of calibration data for new subjects prior to online use. Calibration time can be reduced or eliminated e.g., by subject-to-subject transfer of a pre-trained classifier or unsupervised adaptive classification methods which learn from scratch and adapt over time. While such heuristics work well in practice, none of them can provide theoretical guarantees. Our objective is to modify an event-related potential (ERP) paradigm to work in unison with the machine learning decoder, and thus to achieve a reliable unsupervised calibrationless decoding with a guarantee to recover the true class means. Method We introduce learning from label proportions (LLP) to the BCI community as a new unsupervised, and easy-to-implement classification approach for ERP-based BCIs. The LLP estimates the mean target and non-target responses based on known proportions of these two classes in different groups of the data. We present a visual ERP speller to meet the requirements of LLP. For evaluation, we ran simulations on artificially created data sets and conducted an online BCI study with 13 subjects performing a copy-spelling task. Results Theoretical considerations show that LLP is guaranteed to minimize the loss function similar to a corresponding supervised classifier. LLP performed well in simulations and in the online application, where 84.5% of characters were spelled correctly on average without prior calibration. Significance The continuously adapting LLP classifier is the first unsupervised decoder for ERP BCIs guaranteed to find the optimal decoder. This makes it an ideal solution to avoid tedious calibration sessions. Additionally, LLP works on complementary principles compared to existing unsupervised methods, opening the door for their further enhancement when combined with LLP. PMID:28407016

  2. Best friends' interactions and substance use: The role of friend pressure and unsupervised co-deviancy.

    Science.gov (United States)

    Tsakpinoglou, Florence; Poulin, François

    2017-10-01

    Best friends exert a substantial influence on rising alcohol and marijuana use during adolescence. Two mechanisms occurring within friendship - friend pressure and unsupervised co-deviancy - may partially capture the way friends influence one another. The current study aims to: (1) examine the psychometric properties of a new instrument designed to assess pressure from a youth's best friend and unsupervised co-deviancy; (2) investigate the relative contribution of these processes to alcohol and marijuana use; and (3) determine whether gender moderates these associations. Data were collected through self-report questionnaires completed by 294 Canadian youths (62% female) across two time points (ages 15-16). Principal component analysis yielded a two-factor solution corresponding to friend pressure and unsupervised co-deviancy. Logistic regressions subsequently showed that unsupervised co-deviancy was predictive of an increase in marijuana use one year later. Neither process predicted an increase in alcohol use. Results did not differ as a function of gender. Copyright © 2017 The Foundation for Professionals in Services for Adolescents. Published by Elsevier Ltd. All rights reserved.

  3. A Comparison of Two Approaches to Beta-Flexible Clustering.

    Science.gov (United States)

    Belbin, Lee; And Others

    1992-01-01

    A method for hierarchical agglomerative polythetic (multivariate) clustering, based on unweighted pair group using arithmetic averages (UPGMA) is compared with the original beta-flexible technique, a weighted average method. Reasons the flexible UPGMA strategy is recommended are discussed, focusing on the ability to recover cluster structure over…

  4. Peringkasan Tweet Berdasarkan Trending Topic Twitter Dengan Pembobotan TF-IDF dan Single Linkage AngglomerativeHierarchical Clustering

    Directory of Open Access Journals (Sweden)

    Annisa Annisa

    2016-10-01

    Full Text Available Trending topic is a feature provided by twitter that informs something widely discussed by users in a particular time. The form of a trending topic is a hashtag and can be selected by clicking. However, the number of tweets for each trending topics can be very large, so it will be difficult if we want to know all the contents. So, in order to make easy when reading the topic, a small number of tweets can be selected as the main idea of the topic. In this study, we applied the Agglomerative Single Linkage Hierarchical Clustering by calculating the TF-IDF value for each word in advance. We used 100 trending topics, where each topic consists of 50 tweets in Indonesian. For testing, we provided 30 trending topics which consist of 2 until 9 sub-topics. The result is that each trending topics can be summarized into shorter text contains 2 until 9 tweets. We were able to summarize 1 trending topics exactly same as the topic summarized by human expert. However, the rest of topics corresponded partially with human expert.

  5. Cluster analysis of spontaneous preterm birth phenotypes identifies potential associations among preterm birth mechanisms.

    Science.gov (United States)

    Esplin, M Sean; Manuck, Tracy A; Varner, Michael W; Christensen, Bryce; Biggio, Joseph; Bukowski, Radek; Parry, Samuel; Zhang, Heping; Huang, Hao; Andrews, William; Saade, George; Sadovsky, Yoel; Reddy, Uma M; Ilekis, John

    2015-09-01

    We sought to use an innovative tool that is based on common biologic pathways to identify specific phenotypes among women with spontaneous preterm birth (SPTB) to enhance investigators' ability to identify and to highlight common mechanisms and underlying genetic factors that are responsible for SPTB. We performed a secondary analysis of a prospective case-control multicenter study of SPTB. All cases delivered a preterm singleton at SPTB ≤34.0 weeks' gestation. Each woman was assessed for the presence of underlying SPTB causes. A hierarchic cluster analysis was used to identify groups of women with homogeneous phenotypic profiles. One of the phenotypic clusters was selected for candidate gene association analysis with the use of VEGAS software. One thousand twenty-eight women with SPTB were assigned phenotypes. Hierarchic clustering of the phenotypes revealed 5 major clusters. Cluster 1 (n = 445) was characterized by maternal stress; cluster 2 (n = 294) was characterized by premature membrane rupture; cluster 3 (n = 120) was characterized by familial factors, and cluster 4 (n = 63) was characterized by maternal comorbidities. Cluster 5 (n = 106) was multifactorial and characterized by infection (INF), decidual hemorrhage (DH), and placental dysfunction (PD). These 3 phenotypes were correlated highly by χ(2) analysis (PD and DH, P cluster 3 of SPTB. We identified 5 major clusters of SPTB based on a phenotype tool and hierarch clustering. There was significant correlation between several of the phenotypes. The INS gene was associated with familial factors that were underlying SPTB. Copyright © 2015 Elsevier Inc. All rights reserved.

  6. Hierarchical and Complex System Entropy Clustering Analysis Based Validation for Traditional Chinese Medicine Syndrome Patterns of Chronic Atrophic Gastritis.

    Science.gov (United States)

    Zhang, Yin; Liu, Yue; Li, Yannan; Zhao, Xia; Zhuo, Lin; Zhou, Ajian; Zhang, Li; Su, Zeqi; Chen, Cen; Du, Shiyu; Liu, Daming; Ding, Xia

    2018-03-22

    Chronic atrophic gastritis (CAG) is the precancerous stage of gastric carcinoma. Traditional Chinese Medicine (TCM) has been widely used in treating CAG. This study aimed to reveal core pathogenesis of CAG by validating the TCM syndrome patterns and provide evidence for optimization of treatment strategies. This is a cross-sectional study conducted in 4 hospitals in China. Hierarchical clustering analysis (HCA) and complex system entropy clustering analysis (CSECA) were performed, respectively, to achieve syndrome pattern validation. Based on HCA, 15 common factors were assigned to 6 syndrome patterns: liver depression and spleen deficiency and blood stasis in the stomach collateral, internal harassment of phlegm-heat and blood stasis in the stomach collateral, phlegm-turbidity internal obstruction, spleen yang deficiency, internal harassment of phlegm-heat and spleen deficiency, and spleen qi deficiency. By CSECA, 22 common factors were assigned to 7 syndrome patterns: qi deficiency, qi stagnation, blood stasis, phlegm turbidity, heat, yang deficiency, and yin deficiency. Combination of qi deficiency, qi stagnation, blood stasis, phlegm turbidity, heat, yang deficiency, and yin deficiency may play a crucial role in CAG pathogenesis. In accord with this, treatment strategies by TCM herbal prescriptions should be targeted to regulating qi, activating blood, resolving turbidity, clearing heat, removing toxin, nourishing yin, and warming yang. Further explorations are needed to verify and expand the current conclusions.

  7. Differentiation of Anatolian honey samples from different botanical origins by ATR-FTIR spectroscopy using multivariate analysis.

    Science.gov (United States)

    Gok, Seher; Severcan, Mete; Goormaghtigh, Erik; Kandemir, Irfan; Severcan, Feride

    2015-03-01

    Botanical origin of the nectar predominantly affects the chemical composition of honey. Analytical techniques used for reliable honey authentication are mostly time consuming and expensive. Additionally, they cannot provide 100% efficiency in accurate authentication. Therefore, alternatives for the determination of floral origin of honey need to be developed. This study aims to discriminate characteristic Anatolian honey samples from different botanical origins based on the differences in their molecular content, rather than giving numerical information about the constituents of samples. Another scope of the study is to differentiate inauthentic honey samples from the natural ones precisely. All samples were tested via unsupervised pattern recognition procedures like hierarchical clustering and Principal Component Analysis (PCA). Discrimination of sample groups was achieved successfully with hierarchical clustering over the spectral range of 1800-750 cm(-1) which suggests a good predictive capability of Fourier Transform Infrared (FTIR) spectroscopy and chemometry for the determination of honey floral source. Copyright © 2014 Elsevier Ltd. All rights reserved.

  8. Unsupervised Assessment of Subcutaneous and Visceral Fat by MRI

    DEFF Research Database (Denmark)

    Jørgensen, Peter Stanley; Larsen, Rasmus; Wraae, Kristian

    2009-01-01

    This paper presents a. method for unsupervised assessment of visceral and subcutaneous adipose tissue in the abdominal region by MRI. The identification of the subcutaneous and the visceral regions were achieved by dynamic programming constrained by points acquired from an active shape model...

  9. Intercalating graphene with clusters of Fe3O4 nanocrystals for electrochemical supercapacitors

    Science.gov (United States)

    Ke, Qingqing; Tang, Chunhua; Liu, Yanqiong; Liu, Huajun; Wang, John

    2014-04-01

    A hierarchical nanostructure consisting of graphene sheets intercalated by clusters of Fe3O4 nanocystals is developed for high-performance supercapacitor electrode. Here we show that the negatively charged graphene oxide (GO) and positively charged Fe3O4 clusters enable a strong electrostatic interaction, generating a hierarchical 3D nanostructure, which gives rise to the intercalated composites through a rational hydrothermal process. The electrocapacitive behavior of the resultant composites is systematically investigated by cyclic voltammeter and galvanostatic charge-discharge techniques, where a positive synergistic effect between graphene and Fe3O4 clusters is identified. A maximum specific capacitance of 169 F g-1 is achieved in the Fe3O4 clusters decorated with effectively reduced graphene oxide (Fe3O4-rGO-12h), which is much higher than those of rGO (101 F g-1) and Fe3O4 (68 F g-1) at the current density of 1 Ag-1. Moreover, this intercalated hierarchical nanostructure demonstrates a good capacitance retention, retaining over 88% of the initial capacity after 1000 cycles.

  10. A Multidimensional and Multimembership Clustering Method for Social Networks and Its Application in Customer Relationship Management

    Directory of Open Access Journals (Sweden)

    Peixin Zhao

    2013-01-01

    Full Text Available Community detection in social networks plays an important role in cluster analysis. Many traditional techniques for one-dimensional problems have been proven inadequate for high-dimensional or mixed type datasets due to the data sparseness and attribute redundancy. In this paper we propose a graph-based clustering method for multidimensional datasets. This novel method has two distinguished features: nonbinary hierarchical tree and the multi-membership clusters. The nonbinary hierarchical tree clearly highlights meaningful clusters, while the multimembership feature may provide more useful service strategies. Experimental results on the customer relationship management confirm the effectiveness of the new method.

  11. The Hierarchical Trend Model for property valuation and local price indices

    NARCIS (Netherlands)

    Francke, M.K.; Vos, G.A.

    2002-01-01

    This paper presents a hierarchical trend model (HTM) for selling prices of houses, addressing three main problems: the spatial and temporal dependence of selling prices and the dependency of price index changes on housing quality. In this model the general price trend, cluster-level price trends,

  12. Rotasi Varimax dan Median Hirarki Cluster Pada Program Raskin di Kabupaten Lombok Barat

    Directory of Open Access Journals (Sweden)

    Desy Komalasari

    2015-11-01

    Full Text Available The granting rice program for poor households (Raskin is one of the West Lombok regency government programs for village poverty. The effectiveness of the program relating to 14 criteria for the poor households Raskin recipients (RTS-PM. The 14 criteria have been grouped into several factors using varimax rotation factor analysis, while the RTS-PM have been grouped using hierarchical median cluster analysis. Four factors obtained based on the analysis. First factor was the house existence, the second factor was the financial ability, the third factor was the house existing facilities, and the four factor was the education of the household head and the purchasing power of clothing. The clustering results using hierarchical median cluster analysis formed 3 clusters. The first cluster contains the RTS-PM which have been grouped into first factor; the second cluster contains the RTS-PM which have been grouped into second and third factor; and the third cluster contains the RTS-PM which have been grouped into fourth factor.

  13. Rotasi Varimax dan Median Hirarki Cluster Pada Program Raskin di Kabupaten Lombok Barat

    Directory of Open Access Journals (Sweden)

    Desy Komalasari

    2015-06-01

    Full Text Available The granting rice program for poor households (Raskin is one of the West Lombok regency government programs for village poverty. The effectiveness of the program relating to 14 criteria for the poor households Raskin recipients (RTS-PM. The 14 criteria have been grouped into several factors using varimax rotation factor analysis, while the RTS-PM have been grouped using hierarchical median cluster analysis. Four factors obtained based on the analysis. First factor was the house existence, the second factor was the financial ability, the third factor was the house existing facilities, and the four factor was the education of the household head and the purchasing power of clothing. The clustering results using hierarchical median cluster analysis formed 3 clusters. The first cluster contains the RTS-PM which have been grouped into first factor; the second cluster contains the RTS-PM which have been grouped into second and third factor; and the third cluster contains the RTS-PM which have been grouped into fourth factor.

  14. Unsupervised classification of operator workload from brain signals

    Science.gov (United States)

    Schultze-Kraft, Matthias; Dähne, Sven; Gugler, Manfred; Curio, Gabriel; Blankertz, Benjamin

    2016-06-01

    Objective. In this study we aimed for the classification of operator workload as it is expected in many real-life workplace environments. We explored brain-signal based workload predictors that differ with respect to the level of label information required for training, including entirely unsupervised approaches. Approach. Subjects executed a task on a touch screen that required continuous effort of visual and motor processing with alternating difficulty. We first employed classical approaches for workload state classification that operate on the sensor space of EEG and compared those to the performance of three state-of-the-art spatial filtering methods: common spatial patterns (CSPs) analysis, which requires binary label information; source power co-modulation (SPoC) analysis, which uses the subjects’ error rate as a target function; and canonical SPoC (cSPoC) analysis, which solely makes use of cross-frequency power correlations induced by different states of workload and thus represents an unsupervised approach. Finally, we investigated the effects of fusing brain signals and peripheral physiological measures (PPMs) and examined the added value for improving classification performance. Main results. Mean classification accuracies of 94%, 92% and 82% were achieved with CSP, SPoC, cSPoC, respectively. These methods outperformed the approaches that did not use spatial filtering and they extracted physiologically plausible components. The performance of the unsupervised cSPoC is significantly increased by augmenting it with PPM features. Significance. Our analyses ensured that the signal sources used for classification were of cortical origin and not contaminated with artifacts. Our findings show that workload states can be successfully differentiated from brain signals, even when less and less information from the experimental paradigm is used, thus paving the way for real-world applications in which label information may be noisy or entirely unavailable.

  15. Analog memristive synapse in spiking networks implementing unsupervised learning

    Directory of Open Access Journals (Sweden)

    Erika Covi

    2016-10-01

    Full Text Available Emerging brain-inspired architectures call for devices that can emulate the functionality of biological synapses in order to implement new efficient computational schemes able to solve ill-posed problems. Various devices and solutions are still under investigation and, in this respect, a challenge is opened to the researchers in the field. Indeed, the optimal candidate is a device able to reproduce the complete functionality of a synapse, i.e. the typical synaptic process underlying learning in biological systems (activity-dependent synaptic plasticity. This implies a device able to change its resistance (synaptic strength, or weight upon proper electrical stimuli (synaptic activity and showing several stable resistive states throughout its dynamic range (analog behavior. Moreover, it should be able to perform spike timing dependent plasticity (STDP, an associative homosynaptic plasticity learning rule based on the delay time between the two firing neurons the synapse is connected to. This rule is a fundamental learning protocol in state-of-art networks, because it allows unsupervised learning. Notwithstanding this fact, STDP-based unsupervised learning has been proposed several times mainly for binary synapses rather than multilevel synapses composed of many binary memristors. This paper proposes an HfO2-based analog memristor as a synaptic element which performs STDP within a small spiking neuromorphic network operating unsupervised learning for character recognition. The trained network is able to recognize five characters even in case incomplete or noisy characters are displayed and it is robust to a device-to-device variability of up to +/-30%.

  16. Analog Memristive Synapse in Spiking Networks Implementing Unsupervised Learning.

    Science.gov (United States)

    Covi, Erika; Brivio, Stefano; Serb, Alexander; Prodromakis, Themis; Fanciulli, Marco; Spiga, Sabina

    2016-01-01

    Emerging brain-inspired architectures call for devices that can emulate the functionality of biological synapses in order to implement new efficient computational schemes able to solve ill-posed problems. Various devices and solutions are still under investigation and, in this respect, a challenge is opened to the researchers in the field. Indeed, the optimal candidate is a device able to reproduce the complete functionality of a synapse, i.e., the typical synaptic process underlying learning in biological systems (activity-dependent synaptic plasticity). This implies a device able to change its resistance (synaptic strength, or weight) upon proper electrical stimuli (synaptic activity) and showing several stable resistive states throughout its dynamic range (analog behavior). Moreover, it should be able to perform spike timing dependent plasticity (STDP), an associative homosynaptic plasticity learning rule based on the delay time between the two firing neurons the synapse is connected to. This rule is a fundamental learning protocol in state-of-art networks, because it allows unsupervised learning. Notwithstanding this fact, STDP-based unsupervised learning has been proposed several times mainly for binary synapses rather than multilevel synapses composed of many binary memristors. This paper proposes an HfO 2 -based analog memristor as a synaptic element which performs STDP within a small spiking neuromorphic network operating unsupervised learning for character recognition. The trained network is able to recognize five characters even in case incomplete or noisy images are displayed and it is robust to a device-to-device variability of up to ±30%.

  17. A comparative evaluation of supervised and unsupervised representation learning approaches for anaplastic medulloblastoma differentiation

    Science.gov (United States)

    Cruz-Roa, Angel; Arevalo, John; Basavanhally, Ajay; Madabhushi, Anant; González, Fabio

    2015-01-01

    Learning data representations directly from the data itself is an approach that has shown great success in different pattern recognition problems, outperforming state-of-the-art feature extraction schemes for different tasks in computer vision, speech recognition and natural language processing. Representation learning applies unsupervised and supervised machine learning methods to large amounts of data to find building-blocks that better represent the information in it. Digitized histopathology images represents a very good testbed for representation learning since it involves large amounts of high complex, visual data. This paper presents a comparative evaluation of different supervised and unsupervised representation learning architectures to specifically address open questions on what type of learning architectures (deep or shallow), type of learning (unsupervised or supervised) is optimal. In this paper we limit ourselves to addressing these questions in the context of distinguishing between anaplastic and non-anaplastic medulloblastomas from routine haematoxylin and eosin stained images. The unsupervised approaches evaluated were sparse autoencoders and topographic reconstruct independent component analysis, and the supervised approach was convolutional neural networks. Experimental results show that shallow architectures with more neurons are better than deeper architectures without taking into account local space invariances and that topographic constraints provide useful invariant features in scale and rotations for efficient tumor differentiation.

  18. Comprehensive profiling of DNA methylation in colorectal cancer reveals subgroups with distinct clinicopathological and molecular features

    International Nuclear Information System (INIS)

    Ang, Pei Woon; Soong, Richie; Loh, Marie; Liem, Natalia; Lim, Pei Li; Grieu, Fabienne; Vaithilingam, Aparna; Platell, Cameron; Yong, Wei Peng; Iacopetta, Barry

    2010-01-01

    Most previous studies of the CpG island methylator phenotype (CIMP) in colorectal cancer (CRC) have been conducted on a relatively small numbers of CpG sites. In the present study we performed comprehensive DNA methylation profiling of CRC with the aim of characterizing CIMP subgroups. DNA methylation at 1,505 CpG sites in 807 cancer-related genes was evaluated using the Illumina GoldenGate ® methylation array in 28 normal colonic mucosa and 91 consecutive CRC samples. Methylation data was analyzed using unsupervised hierarchical clustering. CIMP subgroups were compared for various clinicopathological and molecular features including patient age, tumor site, microsatellite instability (MSI), methylation at a consensus panel of CpG islands and mutations in BRAF and KRAS. A total of 202 CpG sites were differentially methylated between tumor and normal tissue. Unsupervised hierarchical clustering of methylation data from these sites revealed the existence of three CRC subgroups referred to as CIMP-low (CIMP-L, 21% of cases), CIMP-mid (CIMP-M, 14%) and CIMP-high (CIMP-H, 65%). In comparison to CIMP-L tumors, CIMP-H tumors were more often located in the proximal colon and showed more frequent mutation of KRAS and BRAF (P < 0.001). Comprehensive DNA methylation profiling identified three CRC subgroups with distinctive clinicopathological and molecular features. This study suggests that both KRAS and BRAF mutations are involved with the CIMP-H pathway of CRC rather than with distinct CIMP subgroups

  19. Comprehensive profiling of DNA methylation in colorectal cancer reveals subgroups with distinct clinicopathological and molecular features

    Directory of Open Access Journals (Sweden)

    Vaithilingam Aparna

    2010-05-01

    Full Text Available Abstract Background Most previous studies of the CpG island methylator phenotype (CIMP in colorectal cancer (CRC have been conducted on a relatively small numbers of CpG sites. In the present study we performed comprehensive DNA methylation profiling of CRC with the aim of characterizing CIMP subgroups. Methods DNA methylation at 1,505 CpG sites in 807 cancer-related genes was evaluated using the Illumina GoldenGate® methylation array in 28 normal colonic mucosa and 91 consecutive CRC samples. Methylation data was analyzed using unsupervised hierarchical clustering. CIMP subgroups were compared for various clinicopathological and molecular features including patient age, tumor site, microsatellite instability (MSI, methylation at a consensus panel of CpG islands and mutations in BRAF and KRAS. Results A total of 202 CpG sites were differentially methylated between tumor and normal tissue. Unsupervised hierarchical clustering of methylation data from these sites revealed the existence of three CRC subgroups referred to as CIMP-low (CIMP-L, 21% of cases, CIMP-mid (CIMP-M, 14% and CIMP-high (CIMP-H, 65%. In comparison to CIMP-L tumors, CIMP-H tumors were more often located in the proximal colon and showed more frequent mutation of KRAS and BRAF (P Conclusions Comprehensive DNA methylation profiling identified three CRC subgroups with distinctive clinicopathological and molecular features. This study suggests that both KRAS and BRAF mutations are involved with the CIMP-H pathway of CRC rather than with distinct CIMP subgroups.

  20. Information-Based Approach to Unsupervised Machine Learning

    Science.gov (United States)

    2013-06-19

    samples with large fitting error. The above optimization problem can be reduced to a quadratic program (Mangasarian & Musicant , 2000), which can be...recognition. Technicheskaya Kibernetica, 3. in Russian. Mallows, C. L. (1973). Some comments on CP . Technometrics, 15, 661–675. Mangasarian, O. L., & Musicant ...to find correspondence between two sets of objects in different domains in an unsupervised way. Photo album summa- rization is a typical application

  1. Extension of mixture-of-experts networks for binary classification of hierarchical data.

    Science.gov (United States)

    Ng, Shu-Kay; McLachlan, Geoffrey J

    2007-09-01

    For many applied problems in the context of medically relevant artificial intelligence, the data collected exhibit a hierarchical or clustered structure. Ignoring the interdependence between hierarchical data can result in misleading classification. In this paper, we extend the mechanism for mixture-of-experts (ME) networks for binary classification of hierarchical data. Another extension is to quantify cluster-specific information on data hierarchy by random effects via the generalized linear mixed-effects model (GLMM). The extension of ME networks is implemented by allowing for correlation in the hierarchical data in both the gating and expert networks via the GLMM. The proposed model is illustrated using a real thyroid disease data set. In our study, we consider 7652 thyroid diagnosis records from 1984 to early 1987 with complete information on 20 attribute values. We obtain 10 independent random splits of the data into a training set and a test set in the proportions 85% and 15%. The test sets are used to assess the generalization performance of the proposed model, based on the percentage of misclassifications. For comparison, the results obtained from the ME network with independence assumption are also included. With the thyroid disease data, the misclassification rate on test sets for the extended ME network is 8.9%, compared to 13.9% for the ME network. In addition, based on model selection methods described in Section 2, a network with two experts is selected. These two expert networks can be considered as modeling two groups of patients with high and low incidence rates. Significant variation among the predicted cluster-specific random effects is detected in the patient group with low incidence rate. It is shown that the extended ME network outperforms the ME network for binary classification of hierarchical data. With the thyroid disease data, useful information on the relative log odds of patients with diagnosed conditions at different periods can be

  2. A new avenue for classification and prediction of olive cultivars using supervised and unsupervised algorithms.

    Directory of Open Access Journals (Sweden)

    Amir H Beiki

    Full Text Available Various methods have been used to identify cultivares of olive trees; herein we used different bioinformatics algorithms to propose new tools to classify 10 cultivares of olive based on RAPD and ISSR genetic markers datasets generated from PCR reactions. Five RAPD markers (OPA0a21, OPD16a, OP01a1, OPD16a1 and OPA0a8 and five ISSR markers (UBC841a4, UBC868a7, UBC841a14, U12BC807a and UBC810a13 selected as the most important markers by all attribute weighting models. K-Medoids unsupervised clustering run on SVM dataset was fully able to cluster each olive cultivar to the right classes. All trees (176 induced by decision tree models generated meaningful trees and UBC841a4 attribute clearly distinguished between foreign and domestic olive cultivars with 100% accuracy. Predictive machine learning algorithms (SVM and Naïve Bayes were also able to predict the right class of olive cultivares with 100% accuracy. For the first time, our results showed data mining techniques can be effectively used to distinguish between plant cultivares and proposed machine learning based systems in this study can predict new olive cultivars with the best possible accuracy.

  3. Using hierarchical clustering of secreted protein families to classify and rank candidate effectors of rust fungi.

    Directory of Open Access Journals (Sweden)

    Diane G O Saunders

    Full Text Available Rust fungi are obligate biotrophic pathogens that cause considerable damage on crop plants. Puccinia graminis f. sp. tritici, the causal agent of wheat stem rust, and Melampsora larici-populina, the poplar leaf rust pathogen, have strong deleterious impacts on wheat and poplar wood production, respectively. Filamentous pathogens such as rust fungi secrete molecules called disease effectors that act as modulators of host cell physiology and can suppress or trigger host immunity. Current knowledge on effectors from other filamentous plant pathogens can be exploited for the characterisation of effectors in the genome of recently sequenced rust fungi. We designed a comprehensive in silico analysis pipeline to identify the putative effector repertoire from the genome of two plant pathogenic rust fungi. The pipeline is based on the observation that known effector proteins from filamentous pathogens have at least one of the following properties: (i contain a secretion signal, (ii are encoded by in planta induced genes, (iii have similarity to haustorial proteins, (iv are small and cysteine rich, (v contain a known effector motif or a nuclear localization signal, (vi are encoded by genes with long intergenic regions, (vii contain internal repeats, and (viii do not contain PFAM domains, except those associated with pathogenicity. We used Markov clustering and hierarchical clustering to classify protein families of rust pathogens and rank them according to their likelihood of being effectors. Using this approach, we identified eight families of candidate effectors that we consider of high value for functional characterization. This study revealed a diverse set of candidate effectors, including families of haustorial expressed secreted proteins and small cysteine-rich proteins. This comprehensive classification of candidate effectors from these devastating rust pathogens is an initial step towards probing plant germplasm for novel resistance components.

  4. Continuous Online Sequence Learning with an Unsupervised Neural Network Model.

    Science.gov (United States)

    Cui, Yuwei; Ahmad, Subutar; Hawkins, Jeff

    2016-09-14

    The ability to recognize and predict temporal sequences of sensory inputs is vital for survival in natural environments. Based on many known properties of cortical neurons, hierarchical temporal memory (HTM) sequence memory recently has been proposed as a theoretical framework for sequence learning in the cortex. In this letter, we analyze properties of HTM sequence memory and apply it to sequence learning and prediction problems with streaming data. We show the model is able to continuously learn a large number of variableorder temporal sequences using an unsupervised Hebbian-like learning rule. The sparse temporal codes formed by the model can robustly handle branching temporal sequences by maintaining multiple predictions until there is sufficient disambiguating evidence. We compare the HTM sequence memory with other sequence learning algorithms, including statistical methods: autoregressive integrated moving average; feedforward neural networks-time delay neural network and online sequential extreme learning machine; and recurrent neural networks-long short-term memory and echo-state networks on sequence prediction problems with both artificial and real-world data. The HTM model achieves comparable accuracy to other state-of-the-art algorithms. The model also exhibits properties that are critical for sequence learning, including continuous online learning, the ability to handle multiple predictions and branching sequences with high-order statistics, robustness to sensor noise and fault tolerance, and good performance without task-specific hyperparameter tuning. Therefore, the HTM sequence memory not only advances our understanding of how the brain may solve the sequence learning problem but is also applicable to real-world sequence learning problems from continuous data streams.

  5. Intraclass Correlation Coefficients in Hierarchical Designs: Evaluation Using Latent Variable Modeling

    Science.gov (United States)

    Raykov, Tenko

    2011-01-01

    Interval estimation of intraclass correlation coefficients in hierarchical designs is discussed within a latent variable modeling framework. A method accomplishing this aim is outlined, which is applicable in two-level studies where participants (or generally lower-order units) are clustered within higher-order units. The procedure can also be…

  6. A method for identifying hierarchical sub-networks / modules and weighting network links based on their similarity in sub-network / module affiliation

    Directory of Open Access Journals (Sweden)

    WenJun Zhang

    2016-06-01

    Full Text Available Some networks, including biological networks, consist of hierarchical sub-networks / modules. Based on my previous study, in present study a method for both identifying hierarchical sub-networks / modules and weighting network links is proposed. It is based on the cluster analysis in which between-node similarity in sets of adjacency nodes is used. Two matrices, linkWeightMat and linkClusterIDs, are achieved by using the algorithm. Two links with both the same weight in linkWeightMat and the same cluster ID in linkClusterIDs belong to the same sub-network / module. Two links with the same weight in linkWeightMat but different cluster IDs in linkClusterIDs belong to two sub-networks / modules at the same hirarchical level. However, a link with an unique cluster ID in linkClusterIDs does not belong to any sub-networks / modules. A sub-network / module of the greater weight is the more connected sub-network / modules. Matlab codes of the algorithm are presented.

  7. Nonlinear dimension reduction and clustering by Minimum Curvilinearity unfold neuropathic pain and tissue embryological classes.

    Science.gov (United States)

    Cannistraci, Carlo Vittorio; Ravasi, Timothy; Montevecchi, Franco Maria; Ideker, Trey; Alessio, Massimo

    2010-09-15

    Nonlinear small datasets, which are characterized by low numbers of samples and very high numbers of measures, occur frequently in computational biology, and pose problems in their investigation. Unsupervised hybrid-two-phase (H2P) procedures-specifically dimension reduction (DR), coupled with clustering-provide valuable assistance, not only for unsupervised data classification, but also for visualization of the patterns hidden in high-dimensional feature space. 'Minimum Curvilinearity' (MC) is a principle that-for small datasets-suggests the approximation of curvilinear sample distances in the feature space by pair-wise distances over their minimum spanning tree (MST), and thus avoids the introduction of any tuning parameter. MC is used to design two novel forms of nonlinear machine learning (NML): Minimum Curvilinear embedding (MCE) for DR, and Minimum Curvilinear affinity propagation (MCAP) for clustering. Compared with several other unsupervised and supervised algorithms, MCE and MCAP, whether individually or combined in H2P, overcome the limits of classical approaches. High performance was attained in the visualization and classification of: (i) pain patients (proteomic measurements) in peripheral neuropathy; (ii) human organ tissues (genomic transcription factor measurements) on the basis of their embryological origin. MC provides a valuable framework to estimate nonlinear distances in small datasets. Its extension to large datasets is prefigured for novel NMLs. Classification of neuropathic pain by proteomic profiles offers new insights for future molecular and systems biology characterization of pain. Improvements in tissue embryological classification refine results obtained in an earlier study, and suggest a possible reinterpretation of skin attribution as mesodermal. https://sites.google.com/site/carlovittoriocannistraci/home.

  8. Unsupervised machine learning account of magnetic transitions in the Hubbard model

    Science.gov (United States)

    Ch'ng, Kelvin; Vazquez, Nick; Khatami, Ehsan

    2018-01-01

    We employ several unsupervised machine learning techniques, including autoencoders, random trees embedding, and t -distributed stochastic neighboring ensemble (t -SNE), to reduce the dimensionality of, and therefore classify, raw (auxiliary) spin configurations generated, through Monte Carlo simulations of small clusters, for the Ising and Fermi-Hubbard models at finite temperatures. Results from a convolutional autoencoder for the three-dimensional Ising model can be shown to produce the magnetization and the susceptibility as a function of temperature with a high degree of accuracy. Quantum fluctuations distort this picture and prevent us from making such connections between the output of the autoencoder and physical observables for the Hubbard model. However, we are able to define an indicator based on the output of the t -SNE algorithm that shows a near perfect agreement with the antiferromagnetic structure factor of the model in two and three spatial dimensions in the weak-coupling regime. t -SNE also predicts a transition to the canted antiferromagnetic phase for the three-dimensional model when a strong magnetic field is present. We show that these techniques cannot be expected to work away from half filling when the "sign problem" in quantum Monte Carlo simulations is present.

  9. A Comparative Evaluation of Unsupervised Anomaly Detection Algorithms for Multivariate Data

    Science.gov (United States)

    Goldstein, Markus; Uchida, Seiichi

    2016-01-01

    Anomaly detection is the process of identifying unexpected items or events in datasets, which differ from the norm. In contrast to standard classification tasks, anomaly detection is often applied on unlabeled data, taking only the internal structure of the dataset into account. This challenge is known as unsupervised anomaly detection and is addressed in many practical applications, for example in network intrusion detection, fraud detection as well as in the life science and medical domain. Dozens of algorithms have been proposed in this area, but unfortunately the research community still lacks a comparative universal evaluation as well as common publicly available datasets. These shortcomings are addressed in this study, where 19 different unsupervised anomaly detection algorithms are evaluated on 10 different datasets from multiple application domains. By publishing the source code and the datasets, this paper aims to be a new well-funded basis for unsupervised anomaly detection research. Additionally, this evaluation reveals the strengths and weaknesses of the different approaches for the first time. Besides the anomaly detection performance, computational effort, the impact of parameter settings as well as the global/local anomaly detection behavior is outlined. As a conclusion, we give an advise on algorithm selection for typical real-world tasks. PMID:27093601

  10. Clustering for data mining a data recovery approach

    CERN Document Server

    Mirkin, Boris

    2005-01-01

    Often considered more as an art than a science, the field of clustering has been dominated by learning through examples and by techniques chosen almost through trial-and-error. Even the most popular clustering methods--K-Means for partitioning the data set and Ward's method for hierarchical clustering--have lacked the theoretical attention that would establish a firm relationship between the two methods and relevant interpretation aids.Rather than the traditional set of ad hoc techniques, Clustering for Data Mining: A Data Recovery Approach presents a theory that not only closes gaps in K-Mean

  11. Accuracy of latent-variable estimation in Bayesian semi-supervised learning.

    Science.gov (United States)

    Yamazaki, Keisuke

    2015-09-01

    Hierarchical probabilistic models, such as Gaussian mixture models, are widely used for unsupervised learning tasks. These models consist of observable and latent variables, which represent the observable data and the underlying data-generation process, respectively. Unsupervised learning tasks, such as cluster analysis, are regarded as estimations of latent variables based on the observable ones. The estimation of latent variables in semi-supervised learning, where some labels are observed, will be more precise than that in unsupervised, and one of the concerns is to clarify the effect of the labeled data. However, there has not been sufficient theoretical analysis of the accuracy of the estimation of latent variables. In a previous study, a distribution-based error function was formulated, and its asymptotic form was calculated for unsupervised learning with generative models. It has been shown that, for the estimation of latent variables, the Bayes method is more accurate than the maximum-likelihood method. The present paper reveals the asymptotic forms of the error function in Bayesian semi-supervised learning for both discriminative and generative models. The results show that the generative model, which uses all of the given data, performs better when the model is well specified. Copyright © 2015 Elsevier Ltd. All rights reserved.

  12. Worldwide clustering of the corruption perception

    Science.gov (United States)

    Paulus, Michal; Kristoufek, Ladislav

    2015-06-01

    We inspect a possible clustering structure of the corruption perception among 134 countries. Using the average linkage clustering, we uncover a well-defined hierarchy in the relationships among countries. Four main clusters are identified and they suggest that countries worldwide can be quite well separated according to their perception of corruption. Moreover, we find a strong connection between corruption levels and a stage of development inside the clusters. The ranking of countries according to their corruption perfectly copies the ranking according to the economic performance measured by the gross domestic product per capita of the member states. To the best of our knowledge, this study is the first one to present an application of hierarchical and clustering methods to the specific case of corruption.

  13. Avoiding Boundary Estimates in Hierarchical Linear Models through Weakly Informative Priors

    Science.gov (United States)

    Chung, Yeojin; Rabe-Hesketh, Sophia; Gelman, Andrew; Dorie, Vincent; Liu, Jinchen

    2012-01-01

    Hierarchical or multilevel linear models are widely used for longitudinal or cross-sectional data on students nested in classes and schools, and are particularly important for estimating treatment effects in cluster-randomized trials, multi-site trials, and meta-analyses. The models can allow for variation in treatment effects, as well as…

  14. Modeling Visit Behaviour in Smart Homes using Unsupervised Learning

    NARCIS (Netherlands)

    Nait Aicha, A.; Englebienne, G.; Kröse, B.

    2014-01-01

    Many algorithms on health monitoring from ambient sensor networks assume that only a single person is present in the home. We present an unsupervised method that models visit behaviour. A Markov modulated multidimensional non-homogeneous Poisson process (M3P2) is described that allows us to model

  15. Re-weighted Discriminatively Embedded K-Means for Multi-view Clustering.

    Science.gov (United States)

    Xu, Jinglin; Han, Junwei; Nie, Feiping; Li, Xuelong

    2017-02-08

    Recent years, more and more multi-view data are widely used in many real world applications. This kind of data (such as image data) are high dimensional and obtained from different feature extractors, which represents distinct perspectives of the data. How to cluster such data efficiently is a challenge. In this paper, we propose a novel multi-view clustering framework, called Re-weighted Discriminatively Embedded KMeans (RDEKM), for this task. The proposed method is a multiview least-absolute residual model which induces robustness to efficiently mitigates the influence of outliers and realizes dimension reduction during multi-view clustering. Specifically, the proposed model is an unsupervised optimization scheme which utilizes Iterative Re-weighted Least Squares to solve leastabsolute residual and adaptively controls the distribution of multiple weights in a re-weighted manner only based on its own low-dimensional subspaces and a common clustering indicator matrix. Furthermore, theoretical analysis (including optimality and convergence analysis) and the optimization algorithm are also presented. Compared to several state-of-the-art multi-view clustering methods, the proposed method substantially improves the accuracy of the clustering results on widely used benchmark datasets, which demonstrates the superiority of the proposed work.

  16. Unsupervised Power Profiling for Mobile Devices

    DEFF Research Database (Denmark)

    Kjærgaard, Mikkel Baun; Blunck, Henrik

    Today, power consumption is a main limitation for mobile phones. To minimize the power consumption of popular and traditionally power-hungry location-based services requires knowledge of how individual phone features consume power, so that those features can be utilized intelligently for optimal...... power savings while at the same time maintaining good quality of service. This paper proposes an unsupervised API-level method for power profiling mobile phones based on genetic algorithms. The method enables accurate profiling of the power consumption of devices and thereby provides the information...

  17. Unsupervised Power Profiling for Mobile Devices

    DEFF Research Database (Denmark)

    Kjærgaard, Mikkel Baun; Blunck, Henrik

    2011-01-01

    Today, power consumption is a main limitation for mobile phones. To minimize the power consumption of popular and traditionally power-hungry location-based services requires knowledge of how individual phone features consume power, so that those features can be utilized intelligently for optimal...... power savings while at the same time maintaining good quality of service. This paper proposes an unsupervised API-level method for power profiling mobile phones based on genetic algorithms. The method enables accurate profiling of the power consumption of devices and thereby provides the information...

  18. Unsupervised information extraction by text segmentation

    CERN Document Server

    Cortez, Eli

    2013-01-01

    A new unsupervised approach to the problem of Information Extraction by Text Segmentation (IETS) is proposed, implemented and evaluated herein. The authors' approach relies on information available on pre-existing data to learn how to associate segments in the input string with attributes of a given domain relying on a very effective set of content-based features. The effectiveness of the content-based features is also exploited to directly learn from test data structure-based features, with no previous human-driven training, a feature unique to the presented approach. Based on the approach, a

  19. Unsupervised detection of salt marsh platforms: a topographic method

    Science.gov (United States)

    Goodwin, Guillaume C. H.; Mudd, Simon M.; Clubb, Fiona J.

    2018-03-01

    Salt marshes filter pollutants, protect coastlines against storm surges, and sequester carbon, yet are under threat from sea level rise and anthropogenic modification. The sustained existence of the salt marsh ecosystem depends on the topographic evolution of marsh platforms. Quantifying marsh platform topography is vital for improving the management of these valuable landscapes. The determination of platform boundaries currently relies on supervised classification methods requiring near-infrared data to detect vegetation, or demands labour-intensive field surveys and digitisation. We propose a novel, unsupervised method to reproducibly isolate salt marsh scarps and platforms from a digital elevation model (DEM), referred to as Topographic Identification of Platforms (TIP). Field observations and numerical models show that salt marshes mature into subhorizontal platforms delineated by subvertical scarps. Based on this premise, we identify scarps as lines of local maxima on a slope raster, then fill landmasses from the scarps upward, thus isolating mature marsh platforms. We test the TIP method using lidar-derived DEMs from six salt marshes in England with varying tidal ranges and geometries, for which topographic platforms were manually isolated from tidal flats. Agreement between manual and unsupervised classification exceeds 94 % for DEM resolutions of 1 m, with all but one site maintaining an accuracy superior to 90 % for resolutions up to 3 m. For resolutions of 1 m, platforms detected with the TIP method are comparable in surface area to digitised platforms and have similar elevation distributions. We also find that our method allows for the accurate detection of local block failures as small as 3 times the DEM resolution. Detailed inspection reveals that although tidal creeks were digitised as part of the marsh platform, unsupervised classification categorises them as part of the tidal flat, causing an increase in false negatives and overall platform

  20. Unsupervised detection of salt marsh platforms: a topographic method

    Directory of Open Access Journals (Sweden)

    G. C. H. Goodwin

    2018-03-01

    Full Text Available Salt marshes filter pollutants, protect coastlines against storm surges, and sequester carbon, yet are under threat from sea level rise and anthropogenic modification. The sustained existence of the salt marsh ecosystem depends on the topographic evolution of marsh platforms. Quantifying marsh platform topography is vital for improving the management of these valuable landscapes. The determination of platform boundaries currently relies on supervised classification methods requiring near-infrared data to detect vegetation, or demands labour-intensive field surveys and digitisation. We propose a novel, unsupervised method to reproducibly isolate salt marsh scarps and platforms from a digital elevation model (DEM, referred to as Topographic Identification of Platforms (TIP. Field observations and numerical models show that salt marshes mature into subhorizontal platforms delineated by subvertical scarps. Based on this premise, we identify scarps as lines of local maxima on a slope raster, then fill landmasses from the scarps upward, thus isolating mature marsh platforms. We test the TIP method using lidar-derived DEMs from six salt marshes in England with varying tidal ranges and geometries, for which topographic platforms were manually isolated from tidal flats. Agreement between manual and unsupervised classification exceeds 94 % for DEM resolutions of 1 m, with all but one site maintaining an accuracy superior to 90 % for resolutions up to 3 m. For resolutions of 1 m, platforms detected with the TIP method are comparable in surface area to digitised platforms and have similar elevation distributions. We also find that our method allows for the accurate detection of local block failures as small as 3 times the DEM resolution. Detailed inspection reveals that although tidal creeks were digitised as part of the marsh platform, unsupervised classification categorises them as part of the tidal flat, causing an increase in false negatives

  1. Knowledge-Based Topic Model for Unsupervised Object Discovery and Localization.

    Science.gov (United States)

    Niu, Zhenxing; Hua, Gang; Wang, Le; Gao, Xinbo

    Unsupervised object discovery and localization is to discover some dominant object classes and localize all of object instances from a given image collection without any supervision. Previous work has attempted to tackle this problem with vanilla topic models, such as latent Dirichlet allocation (LDA). However, in those methods no prior knowledge for the given image collection is exploited to facilitate object discovery. On the other hand, the topic models used in those methods suffer from the topic coherence issue-some inferred topics do not have clear meaning, which limits the final performance of object discovery. In this paper, prior knowledge in terms of the so-called must-links are exploited from Web images on the Internet. Furthermore, a novel knowledge-based topic model, called LDA with mixture of Dirichlet trees, is proposed to incorporate the must-links into topic modeling for object discovery. In particular, to better deal with the polysemy phenomenon of visual words, the must-link is re-defined as that one must-link only constrains one or some topic(s) instead of all topics, which leads to significantly improved topic coherence. Moreover, the must-links are built and grouped with respect to specific object classes, thus the must-links in our approach are semantic-specific , which allows to more efficiently exploit discriminative prior knowledge from Web images. Extensive experiments validated the efficiency of our proposed approach on several data sets. It is shown that our method significantly improves topic coherence and outperforms the unsupervised methods for object discovery and localization. In addition, compared with discriminative methods, the naturally existing object classes in the given image collection can be subtly discovered, which makes our approach well suited for realistic applications of unsupervised object discovery.Unsupervised object discovery and localization is to discover some dominant object classes and localize all of object

  2. Conceptual hierarchical modeling to describe wetland plant community organization

    Science.gov (United States)

    Little, A.M.; Guntenspergen, G.R.; Allen, T.F.H.

    2010-01-01

    Using multivariate analysis, we created a hierarchical modeling process that describes how differently-scaled environmental factors interact to affect wetland-scale plant community organization in a system of small, isolated wetlands on Mount Desert Island, Maine. We followed the procedure: 1) delineate wetland groups using cluster analysis, 2) identify differently scaled environmental gradients using non-metric multidimensional scaling, 3) order gradient hierarchical levels according to spatiotem-poral scale of fluctuation, and 4) assemble hierarchical model using group relationships with ordination axes and post-hoc tests of environmental differences. Using this process, we determined 1) large wetland size and poor surface water chemistry led to the development of shrub fen wetland vegetation, 2) Sphagnum and water chemistry differences affected fen vs. marsh / sedge meadows status within small wetlands, and 3) small-scale hydrologic differences explained transitions between forested vs. non-forested and marsh vs. sedge meadow vegetation. This hierarchical modeling process can help explain how upper level contextual processes constrain biotic community response to lower-level environmental changes. It creates models with more nuanced spatiotemporal complexity than classification and regression tree procedures. Using this process, wetland scientists will be able to generate more generalizable theories of plant community organization, and useful management models. ?? Society of Wetland Scientists 2009.

  3. Manual hierarchical clustering of regional geochemical data using a Bayesian finite mixture model

    International Nuclear Information System (INIS)

    Ellefsen, Karl J.; Smith, David B.

    2016-01-01

    Interpretation of regional scale, multivariate geochemical data is aided by a statistical technique called “clustering.” We investigate a particular clustering procedure by applying it to geochemical data collected in the State of Colorado, United States of America. The clustering procedure partitions the field samples for the entire survey area into two clusters. The field samples in each cluster are partitioned again to create two subclusters, and so on. This manual procedure generates a hierarchy of clusters, and the different levels of the hierarchy show geochemical and geological processes occurring at different spatial scales. Although there are many different clustering methods, we use Bayesian finite mixture modeling with two probability distributions, which yields two clusters. The model parameters are estimated with Hamiltonian Monte Carlo sampling of the posterior probability density function, which usually has multiple modes. Each mode has its own set of model parameters; each set is checked to ensure that it is consistent both with the data and with independent geologic knowledge. The set of model parameters that is most consistent with the independent geologic knowledge is selected for detailed interpretation and partitioning of the field samples. - Highlights: • We evaluate a clustering procedure by applying it to geochemical data. • The procedure generates a hierarchy of clusters. • Different levels of the hierarchy show geochemical processes at different spatial scales. • The clustering method is Bayesian finite mixture modeling. • Model parameters are estimated with Hamiltonian Monte Carlo sampling.

  4. Graph-based unsupervised segmentation algorithm for cultured neuronal networks' structure characterization and modeling.

    Science.gov (United States)

    de Santos-Sierra, Daniel; Sendiña-Nadal, Irene; Leyva, Inmaculada; Almendral, Juan A; Ayali, Amir; Anava, Sarit; Sánchez-Ávila, Carmen; Boccaletti, Stefano

    2015-06-01

    Large scale phase-contrast images taken at high resolution through the life of a cultured neuronal network are analyzed by a graph-based unsupervised segmentation algorithm with a very low computational cost, scaling linearly with the image size. The processing automatically retrieves the whole network structure, an object whose mathematical representation is a matrix in which nodes are identified neurons or neurons' clusters, and links are the reconstructed connections between them. The algorithm is also able to extract any other relevant morphological information characterizing neurons and neurites. More importantly, and at variance with other segmentation methods that require fluorescence imaging from immunocytochemistry techniques, our non invasive measures entitle us to perform a longitudinal analysis during the maturation of a single culture. Such an analysis furnishes the way of individuating the main physical processes underlying the self-organization of the neurons' ensemble into a complex network, and drives the formulation of a phenomenological model yet able to describe qualitatively the overall scenario observed during the culture growth. © 2014 International Society for Advancement of Cytometry.

  5. Enhancement of ELM by Clustering Discrimination Manifold Regularization and Multiobjective FOA for Semisupervised Classification.

    Science.gov (United States)

    Ye, Qing; Pan, Hao; Liu, Changhua

    2015-01-01

    A novel semisupervised extreme learning machine (ELM) with clustering discrimination manifold regularization (CDMR) framework named CDMR-ELM is proposed for semisupervised classification. By using unsupervised fuzzy clustering method, CDMR framework integrates clustering discrimination of both labeled and unlabeled data with twinning constraints regularization. Aiming at further improving the classification accuracy and efficiency, a new multiobjective fruit fly optimization algorithm (MOFOA) is developed to optimize crucial parameters of CDME-ELM. The proposed MOFOA is implemented with two objectives: simultaneously minimizing the number of hidden nodes and mean square error (MSE). The results of experiments on actual datasets show that the proposed semisupervised classifier can obtain better accuracy and efficiency with relatively few hidden nodes compared with other state-of-the-art classifiers.

  6. Enhancement of ELM by Clustering Discrimination Manifold Regularization and Multiobjective FOA for Semisupervised Classification

    Directory of Open Access Journals (Sweden)

    Qing Ye

    2015-01-01

    Full Text Available A novel semisupervised extreme learning machine (ELM with clustering discrimination manifold regularization (CDMR framework named CDMR-ELM is proposed for semisupervised classification. By using unsupervised fuzzy clustering method, CDMR framework integrates clustering discrimination of both labeled and unlabeled data with twinning constraints regularization. Aiming at further improving the classification accuracy and efficiency, a new multiobjective fruit fly optimization algorithm (MOFOA is developed to optimize crucial parameters of CDME-ELM. The proposed MOFOA is implemented with two objectives: simultaneously minimizing the number of hidden nodes and mean square error (MSE. The results of experiments on actual datasets show that the proposed semisupervised classifier can obtain better accuracy and efficiency with relatively few hidden nodes compared with other state-of-the-art classifiers.

  7. Non-Hierarchical Clustering as a method to analyse an open-ended ...

    African Journals Online (AJOL)

    Apple

    Keywords: algebraic thinking; cluster analysis; mathematics education; quantitative analysis. Introduction. Extensive ..... C1, C2 and C3 represent the three centroids of the three clusters formed. .... 6ALd. All these strategies are algebraic and 'high- ... 1995), of the didactical aspects related to teaching .... Brazil, 18-23 July.

  8. Are clusters of dietary patterns and cluster membership stable over time? Results of a longitudinal cluster analysis study.

    Science.gov (United States)

    Walthouwer, Michel Jean Louis; Oenema, Anke; Soetens, Katja; Lechner, Lilian; de Vries, Hein

    2014-11-01

    Developing nutrition education interventions based on clusters of dietary patterns can only be done adequately when it is clear if distinctive clusters of dietary patterns can be derived and reproduced over time, if cluster membership is stable, and if it is predictable which type of people belong to a certain cluster. Hence, this study aimed to: (1) identify clusters of dietary patterns among Dutch adults, (2) test the reproducibility of these clusters and stability of cluster membership over time, and (3) identify sociodemographic predictors of cluster membership and cluster transition. This study had a longitudinal design with online measurements at baseline (N=483) and 6 months follow-up (N=379). Dietary intake was assessed with a validated food frequency questionnaire. A hierarchical cluster analysis was performed, followed by a K-means cluster analysis. Multinomial logistic regression analyses were conducted to identify the sociodemographic predictors of cluster membership and cluster transition. At baseline and follow-up, a comparable three-cluster solution was derived, distinguishing a healthy, moderately healthy, and unhealthy dietary pattern. Male and lower educated participants were significantly more likely to have a less healthy dietary pattern. Further, 251 (66.2%) participants remained in the same cluster, 45 (11.9%) participants changed to an unhealthier cluster, and 83 (21.9%) participants shifted to a healthier cluster. Men and people living alone were significantly more likely to shift toward a less healthy dietary pattern. Distinctive clusters of dietary patterns can be derived. Yet, cluster membership is unstable and only few sociodemographic factors were associated with cluster membership and cluster transition. These findings imply that clusters based on dietary intake may not be suitable as a basis for nutrition education interventions. Copyright © 2014 Elsevier Ltd. All rights reserved.

  9. An unsupervised adaptive strategy for constructing probabilistic roadmaps

    KAUST Repository

    Tapia, L.

    2009-05-01

    Since planning environments are complex and no single planner exists that is best for all problems, much work has been done to explore methods for selecting where and when to apply particular planners. However, these two questions have been difficult to answer, even when adaptive methods meant to facilitate a solution are applied. For example, adaptive solutions such as setting learning rates, hand-classifying spaces, and defining parameters for a library of planners have all been proposed. We demonstrate a strategy based on unsupervised learning methods that makes adaptive planning more practical. The unsupervised strategies require less user intervention, model the topology of the problem in a reasonable and efficient manner, can adapt the sampler depending on characteristics of the problem, and can easily accept new samplers as they become available. Through a series of experiments, we demonstrate that in a wide variety of environments, the regions automatically identified by our technique represent the planning space well both in number and placement.We also show that our technique has little overhead and that it out-performs two existing adaptive methods in all complex cases studied.© 2009 IEEE.

  10. Robustness of Multiple Clustering Algorithms on Hyperspectral Images

    National Research Council Canada - National Science Library

    Williams, Jason P

    2007-01-01

    .... Various clustering algorithms were employed, including a hierarchical method, ISODATA, K-means, and X-means, and were used on a simple two dimensional dataset in order to discover potential problems with the algorithms...

  11. Bilingual Lexical Interactions in an Unsupervised Neural Network Model

    Science.gov (United States)

    Zhao, Xiaowei; Li, Ping

    2010-01-01

    In this paper we present an unsupervised neural network model of bilingual lexical development and interaction. We focus on how the representational structures of the bilingual lexicons can emerge, develop, and interact with each other as a function of the learning history. The results show that: (1) distinct representations for the two lexicons…

  12. Geodesic Flow Kernel Support Vector Machine for Hyperspectral Image Classification by Unsupervised Subspace Feature Transfer

    Directory of Open Access Journals (Sweden)

    Alim Samat

    2016-03-01

    Full Text Available In order to deal with scenarios where the training data, used to deduce a model, and the validation data have different statistical distributions, we study the problem of transformed subspace feature transfer for domain adaptation (DA in the context of hyperspectral image classification via a geodesic Gaussian flow kernel based support vector machine (GFKSVM. To show the superior performance of the proposed approach, conventional support vector machines (SVMs and state-of-the-art DA algorithms, including information-theoretical learning of discriminative cluster for domain adaptation (ITLDC, joint distribution adaptation (JDA, and joint transfer matching (JTM, are also considered. Additionally, unsupervised linear and nonlinear subspace feature transfer techniques including principal component analysis (PCA, randomized nonlinear principal component analysis (rPCA, factor analysis (FA and non-negative matrix factorization (NNMF are investigated and compared. Experiments on two real hyperspectral images show the cross-image classification performances of the GFKSVM, confirming its effectiveness and suitability when applied to hyperspectral images.

  13. Hierarchical structures of correlations networks among Turkey’s exports and imports by currencies

    Science.gov (United States)

    Kocakaplan, Yusuf; Deviren, Bayram; Keskin, Mustafa

    2012-12-01

    We have examined the hierarchical structures of correlations networks among Turkey’s exports and imports by currencies for the 1996-2010 periods, using the concept of a minimal spanning tree (MST) and hierarchical tree (HT) which depend on the concept of ultrametricity. These trees are useful tools for understanding and detecting the global structure, taxonomy and hierarchy in financial markets. We derived a hierarchical organization and build the MSTs and HTs during the 1996-2001 and 2002-2010 periods. The reason for studying two different sub-periods, namely 1996-2001 and 2002-2010, is that the Euro (EUR) came into use in 2001, and some countries have made their exports and imports with Turkey via the EUR since 2002, and in order to test various time-windows and observe temporal evolution. We have carried out bootstrap analysis to associate a value of the statistical reliability to the links of the MSTs and HTs. We have also used the average linkage cluster analysis (ALCA) to observe the cluster structure more clearly. Moreover, we have obtained the bidimensional minimal spanning tree (BMST) due to economic trade being a bidimensional problem. From the structural topologies of these trees, we have identified different clusters of currencies according to their proximity and economic ties. Our results show that some currencies are more important within the network, due to a tighter connection with other currencies. We have also found that the obtained currencies play a key role for Turkey’s exports and imports and have important implications for the design of portfolio and investment strategies.

  14. Unsupervised Data Mining in nanoscale X-ray Spectro-Microscopic Study of NdFeB Magnet.

    Science.gov (United States)

    Duan, Xiaoyue; Yang, Feifei; Antono, Erin; Yang, Wenge; Pianetta, Piero; Ermon, Stefano; Mehta, Apurva; Liu, Yijin

    2016-09-29

    Novel developments in X-ray based spectro-microscopic characterization techniques have increased the rate of acquisition of spatially resolved spectroscopic data by several orders of magnitude over what was possible a few years ago. This accelerated data acquisition, with high spatial resolution at nanoscale and sensitivity to subtle differences in chemistry and atomic structure, provides a unique opportunity to investigate hierarchically complex and structurally heterogeneous systems found in functional devices and materials systems. However, handling and analyzing the large volume data generated poses significant challenges. Here we apply an unsupervised data-mining algorithm known as DBSCAN to study a rare-earth element based permanent magnet material, Nd 2 Fe 14 B. We are able to reduce a large spectro-microscopic dataset of over 300,000 spectra to 3, preserving much of the underlying information. Scientists can easily and quickly analyze in detail three characteristic spectra. Our approach can rapidly provide a concise representation of a large and complex dataset to materials scientists and chemists. For example, it shows that the surface of common Nd 2 Fe 14 B magnet is chemically and structurally very different from the bulk, suggesting a possible surface alteration effect possibly due to the corrosion, which could affect the material's overall properties.

  15. Unsupervised Data Mining in nanoscale X-ray Spectro-Microscopic Study of NdFeB Magnet

    Science.gov (United States)

    Duan, Xiaoyue; Yang, Feifei; Antono, Erin; Yang, Wenge; Pianetta, Piero; Ermon, Stefano; Mehta, Apurva; Liu, Yijin

    2016-09-01

    Novel developments in X-ray based spectro-microscopic characterization techniques have increased the rate of acquisition of spatially resolved spectroscopic data by several orders of magnitude over what was possible a few years ago. This accelerated data acquisition, with high spatial resolution at nanoscale and sensitivity to subtle differences in chemistry and atomic structure, provides a unique opportunity to investigate hierarchically complex and structurally heterogeneous systems found in functional devices and materials systems. However, handling and analyzing the large volume data generated poses significant challenges. Here we apply an unsupervised data-mining algorithm known as DBSCAN to study a rare-earth element based permanent magnet material, Nd2Fe14B. We are able to reduce a large spectro-microscopic dataset of over 300,000 spectra to 3, preserving much of the underlying information. Scientists can easily and quickly analyze in detail three characteristic spectra. Our approach can rapidly provide a concise representation of a large and complex dataset to materials scientists and chemists. For example, it shows that the surface of common Nd2Fe14B magnet is chemically and structurally very different from the bulk, suggesting a possible surface alteration effect possibly due to the corrosion, which could affect the material’s overall properties.

  16. Nonlinear dimension reduction and clustering by Minimum Curvilinearity unfold neuropathic pain and tissue embryological classes

    KAUST Repository

    Cannistraci, Carlo

    2010-09-01

    Motivation: Nonlinear small datasets, which are characterized by low numbers of samples and very high numbers of measures, occur frequently in computational biology, and pose problems in their investigation. Unsupervised hybrid-two-phase (H2P) procedures-specifically dimension reduction (DR), coupled with clustering-provide valuable assistance, not only for unsupervised data classification, but also for visualization of the patterns hidden in high-dimensional feature space. Methods: \\'Minimum Curvilinearity\\' (MC) is a principle that-for small datasets-suggests the approximation of curvilinear sample distances in the feature space by pair-wise distances over their minimum spanning tree (MST), and thus avoids the introduction of any tuning parameter. MC is used to design two novel forms of nonlinear machine learning (NML): Minimum Curvilinear embedding (MCE) for DR, and Minimum Curvilinear affinity propagation (MCAP) for clustering. Results: Compared with several other unsupervised and supervised algorithms, MCE and MCAP, whether individually or combined in H2P, overcome the limits of classical approaches. High performance was attained in the visualization and classification of: (i) pain patients (proteomic measurements) in peripheral neuropathy; (ii) human organ tissues (genomic transcription factor measurements) on the basis of their embryological origin. Conclusion: MC provides a valuable framework to estimate nonlinear distances in small datasets. Its extension to large datasets is prefigured for novel NMLs. Classification of neuropathic pain by proteomic profiles offers new insights for future molecular and systems biology characterization of pain. Improvements in tissue embryological classification refine results obtained in an earlier study, and suggest a possible reinterpretation of skin attribution as mesodermal. © The Author(s) 2010. Published by Oxford University Press.

  17. Remote photoplethysmography system for unsupervised monitoring regional anesthesia effectiveness

    Science.gov (United States)

    Rubins, U.; Miscuks, A.; Marcinkevics, Z.; Lange, M.

    2017-12-01

    Determining the level of regional anesthesia (RA) is vitally important to both an anesthesiologist and surgeon, also knowing the RA level can protect the patient and reduce the time of surgery. Normally to detect the level of RA, usually a simple subjective (sensitivity test) and complicated quantitative methods (thermography, neuromyography, etc.) are used, but there is not yet a standardized method for objective RA detection and evaluation. In this study, the advanced remote photoplethysmography imaging (rPPG) system for unsupervised monitoring of human palm RA is demonstrated. The rPPG system comprises compact video camera with green optical filter, surgical lamp as a light source and a computer with custom-developed software. The algorithm implemented in Matlab software recognizes the palm and two dermatomes (Medial and Ulnar innervation), calculates the perfusion map and perfusion changes in real-time to detect effect of RA. Seven patients (aged 18-80 years) undergoing hand surgery received peripheral nerve brachial plexus blocks during the measurements. Clinical experiments showed that our rPPG system is able to perform unsupervised monitoring of RA.

  18. Hierarchical Spatial Concept Formation Based on Multimodal Information for Human Support Robots.

    Science.gov (United States)

    Hagiwara, Yoshinobu; Inoue, Masakazu; Kobayashi, Hiroyoshi; Taniguchi, Tadahiro

    2018-01-01

    In this paper, we propose a hierarchical spatial concept formation method based on the Bayesian generative model with multimodal information e.g., vision, position and word information. Since humans have the ability to select an appropriate level of abstraction according to the situation and describe their position linguistically, e.g., "I am in my home" and "I am in front of the table," a hierarchical structure of spatial concepts is necessary in order for human support robots to communicate smoothly with users. The proposed method enables a robot to form hierarchical spatial concepts by categorizing multimodal information using hierarchical multimodal latent Dirichlet allocation (hMLDA). Object recognition results using convolutional neural network (CNN), hierarchical k-means clustering result of self-position estimated by Monte Carlo localization (MCL), and a set of location names are used, respectively, as features in vision, position, and word information. Experiments in forming hierarchical spatial concepts and evaluating how the proposed method can predict unobserved location names and position categories are performed using a robot in the real world. Results verify that, relative to comparable baseline methods, the proposed method enables a robot to predict location names and position categories closer to predictions made by humans. As an application example of the proposed method in a home environment, a demonstration in which a human support robot moves to an instructed place based on human speech instructions is achieved based on the formed hierarchical spatial concept.

  19. Hierarchical Spatial Concept Formation Based on Multimodal Information for Human Support Robots

    Directory of Open Access Journals (Sweden)

    Yoshinobu Hagiwara

    2018-03-01

    Full Text Available In this paper, we propose a hierarchical spatial concept formation method based on the Bayesian generative model with multimodal information e.g., vision, position and word information. Since humans have the ability to select an appropriate level of abstraction according to the situation and describe their position linguistically, e.g., “I am in my home” and “I am in front of the table,” a hierarchical structure of spatial concepts is necessary in order for human support robots to communicate smoothly with users. The proposed method enables a robot to form hierarchical spatial concepts by categorizing multimodal information using hierarchical multimodal latent Dirichlet allocation (hMLDA. Object recognition results using convolutional neural network (CNN, hierarchical k-means clustering result of self-position estimated by Monte Carlo localization (MCL, and a set of location names are used, respectively, as features in vision, position, and word information. Experiments in forming hierarchical spatial concepts and evaluating how the proposed method can predict unobserved location names and position categories are performed using a robot in the real world. Results verify that, relative to comparable baseline methods, the proposed method enables a robot to predict location names and position categories closer to predictions made by humans. As an application example of the proposed method in a home environment, a demonstration in which a human support robot moves to an instructed place based on human speech instructions is achieved based on the formed hierarchical spatial concept.

  20. Unsupervised progressive elastic band exercises for frail geriatric inpatients objectively monitored by new exercise-integrated technology

    DEFF Research Database (Denmark)

    Rathleff, Camilla Rams; Bandholm, T.; Spaich, Erika Geraldina

    2017-01-01

    the amount of supervised training, and unsupervised training could possibly supplement supervised training thereby increasing the total exercise dose during admission. A new valid and reliable technology, the BandCizer, objectively measures the exact training dosage performed. The purpose was to investigate...... feasibility and acceptability of an unsupervised progressive strength training intervention monitored by BandCizer for frail geriatric inpatients. Methods: This feasibility trial included 15 frail inpatients at a geriatric ward. At hospitalization, the patients were prescribed two elastic band exercises...... of 2-min pauses and a time-under-tension of 8 s. The feasibility criterion for the unsupervised progressive exercises was that 33% of the recommended number of sets would be performed by at least 30% of patients. In addition, patients and staff were interviewed about their experiences...

  1. Clustering and Bayesian hierarchical modeling for the definition of informative prior distributions in hydrogeology

    Science.gov (United States)

    Cucchi, K.; Kawa, N.; Hesse, F.; Rubin, Y.

    2017-12-01

    In order to reduce uncertainty in the prediction of subsurface flow and transport processes, practitioners should use all data available. However, classic inverse modeling frameworks typically only make use of information contained in in-situ field measurements to provide estimates of hydrogeological parameters. Such hydrogeological information about an aquifer is difficult and costly to acquire. In this data-scarce context, the transfer of ex-situ information coming from previously investigated sites can be critical for improving predictions by better constraining the estimation procedure. Bayesian inverse modeling provides a coherent framework to represent such ex-situ information by virtue of the prior distribution and combine them with in-situ information from the target site. In this study, we present an innovative data-driven approach for defining such informative priors for hydrogeological parameters at the target site. Our approach consists in two steps, both relying on statistical and machine learning methods. The first step is data selection; it consists in selecting sites similar to the target site. We use clustering methods for selecting similar sites based on observable hydrogeological features. The second step is data assimilation; it consists in assimilating data from the selected similar sites into the informative prior. We use a Bayesian hierarchical model to account for inter-site variability and to allow for the assimilation of multiple types of site-specific data. We present the application and validation of the presented methods on an established database of hydrogeological parameters. Data and methods are implemented in the form of an open-source R-package and therefore facilitate easy use by other practitioners.

  2. Clusters of Galaxies

    Science.gov (United States)

    Huchtmeier, W. K.; Richter, O. G.; Materne, J.

    1981-09-01

    The large-scale structure of the universe is dominated by clustering. Most galaxies seem to be members of pairs, groups, clusters, and superclusters. To that degree we are able to recognize a hierarchical structure of the universe. Our local group of galaxies (LG) is centred on two large spiral galaxies: the Andromeda nebula and our own galaxy. Three sr:naller galaxies - like M 33 - and at least 23 dwarf galaxies (KraanKorteweg and Tammann, 1979, Astronomische Nachrichten, 300, 181) can be found in the evironment of these two large galaxies. Neighbouring groups have comparable sizes (about 1 Mpc in extent) and comparable numbers of bright members. Small dwarf galaxies cannot at present be observed at great distances.

  3. Automated lesion detection on MRI scans using combined unsupervised and supervised methods

    International Nuclear Information System (INIS)

    Guo, Dazhou; Fridriksson, Julius; Fillmore, Paul; Rorden, Christopher; Yu, Hongkai; Zheng, Kang; Wang, Song

    2015-01-01

    Accurate and precise detection of brain lesions on MR images (MRI) is paramount for accurately relating lesion location to impaired behavior. In this paper, we present a novel method to automatically detect brain lesions from a T1-weighted 3D MRI. The proposed method combines the advantages of both unsupervised and supervised methods. First, unsupervised methods perform a unified segmentation normalization to warp images from the native space into a standard space and to generate probability maps for different tissue types, e.g., gray matter, white matter and fluid. This allows us to construct an initial lesion probability map by comparing the normalized MRI to healthy control subjects. Then, we perform non-rigid and reversible atlas-based registration to refine the probability maps of gray matter, white matter, external CSF, ventricle, and lesions. These probability maps are combined with the normalized MRI to construct three types of features, with which we use supervised methods to train three support vector machine (SVM) classifiers for a combined classifier. Finally, the combined classifier is used to accomplish lesion detection. We tested this method using T1-weighted MRIs from 60 in-house stroke patients. Using leave-one-out cross validation, the proposed method can achieve an average Dice coefficient of 73.1 % when compared to lesion maps hand-delineated by trained neurologists. Furthermore, we tested the proposed method on the T1-weighted MRIs in the MICCAI BRATS 2012 dataset. The proposed method can achieve an average Dice coefficient of 66.5 % in comparison to the expert annotated tumor maps provided in MICCAI BRATS 2012 dataset. In addition, on these two test datasets, the proposed method shows competitive performance to three state-of-the-art methods, including Stamatakis et al., Seghier et al., and Sanjuan et al. In this paper, we introduced a novel automated procedure for lesion detection from T1-weighted MRIs by combining both an unsupervised and a

  4. MULTI-K: accurate classification of microarray subtypes using ensemble k-means clustering

    Directory of Open Access Journals (Sweden)

    Ashlock Daniel

    2009-08-01

    Full Text Available Abstract Background Uncovering subtypes of disease from microarray samples has important clinical implications such as survival time and sensitivity of individual patients to specific therapies. Unsupervised clustering methods have been used to classify this type of data. However, most existing methods focus on clusters with compact shapes and do not reflect the geometric complexity of the high dimensional microarray clusters, which limits their performance. Results We present a cluster-number-based ensemble clustering algorithm, called MULTI-K, for microarray sample classification, which demonstrates remarkable accuracy. The method amalgamates multiple k-means runs by varying the number of clusters and identifies clusters that manifest the most robust co-memberships of elements. In addition to the original algorithm, we newly devised the entropy-plot to control the separation of singletons or small clusters. MULTI-K, unlike the simple k-means or other widely used methods, was able to capture clusters with complex and high-dimensional structures accurately. MULTI-K outperformed other methods including a recently developed ensemble clustering algorithm in tests with five simulated and eight real gene-expression data sets. Conclusion The geometric complexity of clusters should be taken into account for accurate classification of microarray data, and ensemble clustering applied to the number of clusters tackles the problem very well. The C++ code and the data sets tested are available from the authors.

  5. MULTI-K: accurate classification of microarray subtypes using ensemble k-means clustering.

    Science.gov (United States)

    Kim, Eun-Youn; Kim, Seon-Young; Ashlock, Daniel; Nam, Dougu

    2009-08-22

    Uncovering subtypes of disease from microarray samples has important clinical implications such as survival time and sensitivity of individual patients to specific therapies. Unsupervised clustering methods have been used to classify this type of data. However, most existing methods focus on clusters with compact shapes and do not reflect the geometric complexity of the high dimensional microarray clusters, which limits their performance. We present a cluster-number-based ensemble clustering algorithm, called MULTI-K, for microarray sample classification, which demonstrates remarkable accuracy. The method amalgamates multiple k-means runs by varying the number of clusters and identifies clusters that manifest the most robust co-memberships of elements. In addition to the original algorithm, we newly devised the entropy-plot to control the separation of singletons or small clusters. MULTI-K, unlike the simple k-means or other widely used methods, was able to capture clusters with complex and high-dimensional structures accurately. MULTI-K outperformed other methods including a recently developed ensemble clustering algorithm in tests with five simulated and eight real gene-expression data sets. The geometric complexity of clusters should be taken into account for accurate classification of microarray data, and ensemble clustering applied to the number of clusters tackles the problem very well. The C++ code and the data sets tested are available from the authors.

  6. Short-Term Wind Power Forecasting Based on Clustering Pre-Calculated CFD Method

    Directory of Open Access Journals (Sweden)

    Yimei Wang

    2018-04-01

    Full Text Available To meet the increasing wind power forecasting (WPF demands of newly built wind farms without historical data, physical WPF methods are widely used. The computational fluid dynamics (CFD pre-calculated flow fields (CPFF-based WPF is a promising physical approach, which can balance well the competing demands of computational efficiency and accuracy. To enhance its adaptability for wind farms in complex terrain, a WPF method combining wind turbine clustering with CPFF is first proposed where the wind turbines in the wind farm are clustered and a forecasting is undertaken for each cluster. K-means, hierarchical agglomerative and spectral analysis methods are used to establish the wind turbine clustering models. The Silhouette Coefficient, Calinski-Harabaz index and within-between index are proposed as criteria to evaluate the effectiveness of the established clustering models. Based on different clustering methods and schemes, various clustering databases are built for clustering pre-calculated CFD (CPCC-based short-term WPF. For the wind farm case studied, clustering evaluation criteria show that hierarchical agglomerative clustering has reasonable results, spectral clustering is better and K-means gives the best performance. The WPF results produced by different clustering databases also prove the effectiveness of the three evaluation criteria in turn. The newly developed CPCC model has a much higher WPF accuracy than the CPFF model without using clustering techniques, both on temporal and spatial scales. The research provides supports for both the development and improvement of short-term physical WPF systems.

  7. OMERACT-based fibromyalgia symptom subgroups: an exploratory cluster analysis.

    Science.gov (United States)

    Vincent, Ann; Hoskin, Tanya L; Whipple, Mary O; Clauw, Daniel J; Barton, Debra L; Benzo, Roberto P; Williams, David A

    2014-10-16

    The aim of this study was to identify subsets of patients with fibromyalgia with similar symptom profiles using the Outcome Measures in Rheumatology (OMERACT) core symptom domains. Female patients with a diagnosis of fibromyalgia and currently meeting fibromyalgia research survey criteria completed the Brief Pain Inventory, the 30-item Profile of Mood States, the Medical Outcomes Sleep Scale, the Multidimensional Fatigue Inventory, the Multiple Ability Self-Report Questionnaire, the Fibromyalgia Impact Questionnaire-Revised (FIQ-R) and the Short Form-36 between 1 June 2011 and 31 October 2011. Hierarchical agglomerative clustering was used to identify subgroups of patients with similar symptom profiles. To validate the results from this sample, hierarchical agglomerative clustering was repeated in an external sample of female patients with fibromyalgia with similar inclusion criteria. A total of 581 females with a mean age of 55.1 (range, 20.1 to 90.2) years were included. A four-cluster solution best fit the data, and each clustering variable differed significantly (P FIQ-R total scores (P = 0.0004)). In our study, we incorporated core OMERACT symptom domains, which allowed for clustering based on a comprehensive symptom profile. Although our exploratory cluster solution needs confirmation in a longitudinal study, this approach could provide a rationale to support the study of individualized clinical evaluation and intervention.

  8. PosQ: Unsupervised Fingerprinting and Visualization of GPS Positioning Quality

    DEFF Research Database (Denmark)

    Kjærgaard, Mikkel Baun; Weckemann, Kay

    . This paper proposes PosQ, a system for unsupervised fingerprinting and visualization of GPS positioning quality. PosQ provides quality maps to position-based applications and visual overlays to users and managers to reveal the positioning quality in a local environment. The system reveals the quality both...

  9. A Novel Unsupervised Adaptive Learning Method for Long-Term Electromyography (EMG) Pattern Recognition

    Science.gov (United States)

    Huang, Qi; Yang, Dapeng; Jiang, Li; Zhang, Huajie; Liu, Hong; Kotani, Kiyoshi

    2017-01-01

    Performance degradation will be caused by a variety of interfering factors for pattern recognition-based myoelectric control methods in the long term. This paper proposes an adaptive learning method with low computational cost to mitigate the effect in unsupervised adaptive learning scenarios. We presents a particle adaptive classifier (PAC), by constructing a particle adaptive learning strategy and universal incremental least square support vector classifier (LS-SVC). We compared PAC performance with incremental support vector classifier (ISVC) and non-adapting SVC (NSVC) in a long-term pattern recognition task in both unsupervised and supervised adaptive learning scenarios. Retraining time cost and recognition accuracy were compared by validating the classification performance on both simulated and realistic long-term EMG data. The classification results of realistic long-term EMG data showed that the PAC significantly decreased the performance degradation in unsupervised adaptive learning scenarios compared with NSVC (9.03% ± 2.23%, p < 0.05) and ISVC (13.38% ± 2.62%, p = 0.001), and reduced the retraining time cost compared with ISVC (2 ms per updating cycle vs. 50 ms per updating cycle). PMID:28608824

  10. A Novel Unsupervised Adaptive Learning Method for Long-Term Electromyography (EMG Pattern Recognition

    Directory of Open Access Journals (Sweden)

    Qi Huang

    2017-06-01

    Full Text Available Performance degradation will be caused by a variety of interfering factors for pattern recognition-based myoelectric control methods in the long term. This paper proposes an adaptive learning method with low computational cost to mitigate the effect in unsupervised adaptive learning scenarios. We presents a particle adaptive classifier (PAC, by constructing a particle adaptive learning strategy and universal incremental least square support vector classifier (LS-SVC. We compared PAC performance with incremental support vector classifier (ISVC and non-adapting SVC (NSVC in a long-term pattern recognition task in both unsupervised and supervised adaptive learning scenarios. Retraining time cost and recognition accuracy were compared by validating the classification performance on both simulated and realistic long-term EMG data. The classification results of realistic long-term EMG data showed that the PAC significantly decreased the performance degradation in unsupervised adaptive learning scenarios compared with NSVC (9.03% ± 2.23%, p < 0.05 and ISVC (13.38% ± 2.62%, p = 0.001, and reduced the retraining time cost compared with ISVC (2 ms per updating cycle vs. 50 ms per updating cycle.

  11. Exploring the individual patterns of spiritual well-being in people newly diagnosed with advanced cancer: a cluster analysis.

    Science.gov (United States)

    Bai, Mei; Dixon, Jane; Williams, Anna-Leila; Jeon, Sangchoon; Lazenby, Mark; McCorkle, Ruth

    2016-11-01

    Research shows that spiritual well-being correlates positively with quality of life (QOL) for people with cancer, whereas contradictory findings are frequently reported with respect to the differentiated associations between dimensions of spiritual well-being, namely peace, meaning and faith, and QOL. This study aimed to examine individual patterns of spiritual well-being among patients newly diagnosed with advanced cancer. Cluster analysis was based on the twelve items of the 12-item Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Scale at Time 1. A combination of hierarchical and k-means (non-hierarchical) clustering methods was employed to jointly determine the number of clusters. Self-rated health, depressive symptoms, peace, meaning and faith, and overall QOL were compared at Time 1 and Time 2. Hierarchical and k-means clustering methods both suggested four clusters. Comparison of the four clusters supported statistically significant and clinically meaningful differences in QOL outcomes among clusters while revealing contrasting relations of faith with QOL. Cluster 1, Cluster 3, and Cluster 4 represented high, medium, and low levels of overall QOL, respectively, with correspondingly high, medium, and low levels of peace, meaning, and faith. Cluster 2 was distinguished from other clusters by its medium levels of overall QOL, peace, and meaning and low level of faith. This study provides empirical support for individual difference in response to a newly diagnosed cancer and brings into focus conceptual and methodological challenges associated with the measure of spiritual well-being, which may partly contribute to the attenuated relation between faith and QOL.

  12. Genome-wide expression patterns associated with oncogenesis and sarcomatous transdifferentation of cholangiocarcinoma

    International Nuclear Information System (INIS)

    Seol, Min-A; Kim, Dae-Ghon; Chu, In-Sun; Lee, Mi-Jin; Yu, Goung-Ran; Cui, Xiang-Dan; Cho, Baik-Hwan; Ahn, Eun-Kyung; Leem, Sun-Hee; Kim, In-Hee

    2011-01-01

    The molecular mechanisms of CC (cholangiocarcinoma) oncogenesis and progression are poorly understood. This study aimed to determine the genome-wide expression of genes related to CC oncogenesis and sarcomatous transdifferentiation. Genes that were differentially expressed between CC cell lines or tissues and cultured normal biliary epithelial (NBE) cells were identified using DNA microarray technology. Expressions were validated in human CC tissues and cells. Using unsupervised hierarchical clustering analysis of the cell line and tissue samples, we identified a set of 342 commonly regulated (>2-fold change) genes. Of these, 53, including tumor-related genes, were upregulated, and 289, including tumor suppressor genes, were downregulated (<0.5 fold change). Expression of SPP1, EFNB2, E2F2, IRX3, PTTG1, PPARγ, KRT17, UCHL1, IGFBP7 and SPARC proteins was immunohistochemically verified in human and hamster CC tissues. Additional unsupervised hierarchical clustering analysis of sarcomatoid CC cells compared to three adenocarcinomatous CC cell lines revealed 292 differentially upregulated genes (>4-fold change), and 267 differentially downregulated genes (<0.25 fold change). The expression of 12 proteins was validated in the CC cell lines by immunoblot analysis and immunohistochemical staining. Of the proteins analyzed, we found upregulation of the expression of the epithelial-mesenchymal transition (EMT)-related proteins VIM and TWIST1, and restoration of the methylation-silenced proteins LDHB, BNIP3, UCHL1, and NPTX2 during sarcomatoid transdifferentiation of CC. The deregulation of oncogenes, tumor suppressor genes, and methylation-related genes may be useful in identifying molecular targets for CC diagnosis and prognosis

  13. Clustering for Different Scales of Measurement - the Gap-Ratio Weighted K-means Algorithm

    OpenAIRE

    Guérin, Joris; Gibaru, Olivier; Thiery, Stéphane; Nyiri, Eric

    2017-01-01

    This paper describes a method for clustering data that are spread out over large regions and which dimensions are on different scales of measurement. Such an algorithm was developed to implement a robotics application consisting in sorting and storing objects in an unsupervised way. The toy dataset used to validate such application consists of Lego bricks of different shapes and colors. The uncontrolled lighting conditions together with the use of RGB color features, respectively involve data...

  14. Not all stars form in clusters - measuring the kinematics of OB associations with Gaia

    Science.gov (United States)

    Ward, Jacob L.; Kruijssen, J. M. Diederik

    2018-04-01

    It is often stated that star clusters are the fundamental units of star formation and that most (if not all) stars form in dense stellar clusters. In this monolithic formation scenario, low-density OB associations are formed from the expansion of gravitationally bound clusters following gas expulsion due to stellar feedback. N-body simulations of this process show that OB associations formed this way retain signs of expansion and elevated radial anisotropy over tens of Myr. However, recent theoretical and observational studies suggest that star formation is a hierarchical process, following the fractal nature of natal molecular clouds and allowing the formation of large-scale associations in situ. We distinguish between these two scenarios by characterizing the kinematics of OB associations using the Tycho-Gaia Astrometric Solution catalogue. To this end, we quantify four key kinematic diagnostics: the number ratio of stars with positive radial velocities to those with negative radial velocities, the median radial velocity, the median radial velocity normalized by the tangential velocity, and the radial anisotropy parameter. Each quantity presents a useful diagnostic of whether the association was more compact in the past. We compare these diagnostics to models representing random motion and the expanding products of monolithic cluster formation. None of these diagnostics show evidence of expansion, either from a single cluster or multiple clusters, and the observed kinematics are better represented by a random velocity distribution. This result favours the hierarchical star formation model in which a minority of stars forms in bound clusters and large-scale, hierarchically structured associations are formed in situ.

  15. Supervised and Unsupervised Speaker Adaptation in the NIST 2005 Speaker Recognition Evaluation

    National Research Council Canada - National Science Library

    Hansen, Eric G; Slyh, Raymond E; Anderson, Timothy R

    2006-01-01

    Starting in 2004, the annual NIST Speaker Recognition Evaluation (SRE) has added an optional unsupervised speaker adaptation track where test files are processed sequentially and one may update the target model...

  16. Hierarchical Cluster Analysis of Three-Dimensional Reconstructions of Unbiased Sampled Microglia Shows not Continuous Morphological Changes from Stage 1 to 2 after Multiple Dengue Infections in Callithrix penicillata

    Science.gov (United States)

    Diniz, Daniel G.; Silva, Geane O.; Naves, Thaís B.; Fernandes, Taiany N.; Araújo, Sanderson C.; Diniz, José A. P.; de Farias, Luis H. S.; Sosthenes, Marcia C. K.; Diniz, Cristovam G.; Anthony, Daniel C.; da Costa Vasconcelos, Pedro F.; Picanço Diniz, Cristovam W.

    2016-01-01

    It is known that microglial morphology and function are related, but few studies have explored the subtleties of microglial morphological changes in response to specific pathogens. In the present report we quantitated microglia morphological changes in a monkey model of dengue disease with virus CNS invasion. To mimic multiple infections that usually occur in endemic areas, where higher dengue infection incidence and abundant mosquito vectors carrying different serotypes coexist, subjects received once a week subcutaneous injections of DENV3 (genotype III)-infected culture supernatant followed 24 h later by an injection of anti-DENV2 antibody. Control animals received either weekly anti-DENV2 antibodies, or no injections. Brain sections were immunolabeled for DENV3 antigens and IBA-1. Random and systematic microglial samples were taken from the polymorphic layer of dentate gyrus for 3-D reconstructions, where we found intense immunostaining for TNFα and DENV3 virus antigens. We submitted all bi- or multimodal morphological parameters of microglia to hierarchical cluster analysis and found two major morphological phenotypes designated types I and II. Compared to type I (stage 1), type II microglia were more complex; displaying higher number of nodes, processes and trees and larger surface area and volumes (stage 2). Type II microglia were found only in infected monkeys, whereas type I microglia was found in both control and infected subjects. Hierarchical cluster analysis of morphological parameters of 3-D reconstructions of random and systematic selected samples in control and ADE dengue infected monkeys suggests that microglia morphological changes from stage 1 to stage 2 may not be continuous. PMID:27047345

  17. Agglomerative concentric hypersphere clustering applied to structural damage detection

    Science.gov (United States)

    Silva, Moisés; Santos, Adam; Santos, Reginaldo; Figueiredo, Eloi; Sales, Claudomiro; Costa, João C. W. A.

    2017-08-01

    The present paper proposes a novel cluster-based method, named as agglomerative concentric hypersphere (ACH), to detect structural damage in engineering structures. Continuous structural monitoring systems often require unsupervised approaches to automatically infer the health condition of a structure. However, when a structure is under linear and nonlinear effects caused by environmental and operational variability, data normalization procedures are also required to overcome these effects. The proposed approach aims, through a straightforward clustering procedure, to discover automatically the optimal number of clusters, representing the main state conditions of a structural system. Three initialization procedures are introduced to evaluate the impact of deterministic and stochastic initializations on the performance of this approach. The ACH is compared to state-of-the-art approaches, based on Gaussian mixture models and Mahalanobis squared distance, on standard data sets from a post-tensioned bridge located in Switzerland: the Z-24 Bridge. The proposed approach demonstrates more efficiency in modeling the normal condition of the structure and its corresponding main clusters. Furthermore, it reveals a better classification performance than the alternative ones in terms of false-positive and false-negative indications of damage, demonstrating a promising applicability in real-world structural health monitoring scenarios.

  18. Deep supervised, but not unsupervised, models may explain IT cortical representation.

    Directory of Open Access Journals (Sweden)

    Seyed-Mahdi Khaligh-Razavi

    2014-11-01

    Full Text Available Inferior temporal (IT cortex in human and nonhuman primates serves visual object recognition. Computational object-vision models, although continually improving, do not yet reach human performance. It is unclear to what extent the internal representations of computational models can explain the IT representation. Here we investigate a wide range of computational model representations (37 in total, testing their categorization performance and their ability to account for the IT representational geometry. The models include well-known neuroscientific object-recognition models (e.g. HMAX, VisNet along with several models from computer vision (e.g. SIFT, GIST, self-similarity features, and a deep convolutional neural network. We compared the representational dissimilarity matrices (RDMs of the model representations with the RDMs obtained from human IT (measured with fMRI and monkey IT (measured with cell recording for the same set of stimuli (not used in training the models. Better performing models were more similar to IT in that they showed greater clustering of representational patterns by category. In addition, better performing models also more strongly resembled IT in terms of their within-category representational dissimilarities. Representational geometries were significantly correlated between IT and many of the models. However, the categorical clustering observed in IT was largely unexplained by the unsupervised models. The deep convolutional network, which was trained by supervision with over a million category-labeled images, reached the highest categorization performance and also best explained IT, although it did not fully explain the IT data. Combining the features of this model with appropriate weights and adding linear combinations that maximize the margin between animate and inanimate objects and between faces and other objects yielded a representation that fully explained our IT data. Overall, our results suggest that explaining

  19. Teacher and learner: Supervised and unsupervised learning in communities.

    Science.gov (United States)

    Shafto, Michael G; Seifert, Colleen M

    2015-01-01

    How far can teaching methods go to enhance learning? Optimal methods of teaching have been considered in research on supervised and unsupervised learning. Locally optimal methods are usually hybrids of teaching and self-directed approaches. The costs and benefits of specific methods have been shown to depend on the structure of the learning task, the learners, the teachers, and the environment.

  20. Flexible manifold embedding: a framework for semi-supervised and unsupervised dimension reduction.

    Science.gov (United States)

    Nie, Feiping; Xu, Dong; Tsang, Ivor Wai-Hung; Zhang, Changshui

    2010-07-01

    We propose a unified manifold learning framework for semi-supervised and unsupervised dimension reduction by employing a simple but effective linear regression function to map the new data points. For semi-supervised dimension reduction, we aim to find the optimal prediction labels F for all the training samples X, the linear regression function h(X) and the regression residue F(0) = F - h(X) simultaneously. Our new objective function integrates two terms related to label fitness and manifold smoothness as well as a flexible penalty term defined on the residue F(0). Our Semi-Supervised learning framework, referred to as flexible manifold embedding (FME), can effectively utilize label information from labeled data as well as a manifold structure from both labeled and unlabeled data. By modeling the mismatch between h(X) and F, we show that FME relaxes the hard linear constraint F = h(X) in manifold regularization (MR), making it better cope with the data sampled from a nonlinear manifold. In addition, we propose a simplified version (referred to as FME/U) for unsupervised dimension reduction. We also show that our proposed framework provides a unified view to explain and understand many semi-supervised, supervised and unsupervised dimension reduction techniques. Comprehensive experiments on several benchmark databases demonstrate the significant improvement over existing dimension reduction algorithms.

  1. Bias correction in the hierarchical likelihood approach to the analysis of multivariate survival data.

    Science.gov (United States)

    Jeon, Jihyoun; Hsu, Li; Gorfine, Malka

    2012-07-01

    Frailty models are useful for measuring unobserved heterogeneity in risk of failures across clusters, providing cluster-specific risk prediction. In a frailty model, the latent frailties shared by members within a cluster are assumed to act multiplicatively on the hazard function. In order to obtain parameter and frailty variate estimates, we consider the hierarchical likelihood (H-likelihood) approach (Ha, Lee and Song, 2001. Hierarchical-likelihood approach for frailty models. Biometrika 88, 233-243) in which the latent frailties are treated as "parameters" and estimated jointly with other parameters of interest. We find that the H-likelihood estimators perform well when the censoring rate is low, however, they are substantially biased when the censoring rate is moderate to high. In this paper, we propose a simple and easy-to-implement bias correction method for the H-likelihood estimators under a shared frailty model. We also extend the method to a multivariate frailty model, which incorporates complex dependence structure within clusters. We conduct an extensive simulation study and show that the proposed approach performs very well for censoring rates as high as 80%. We also illustrate the method with a breast cancer data set. Since the H-likelihood is the same as the penalized likelihood function, the proposed bias correction method is also applicable to the penalized likelihood estimators.

  2. Deep Patient: An Unsupervised Representation to Predict the Future of Patients from the Electronic Health Records

    Science.gov (United States)

    Miotto, Riccardo; Li, Li; Kidd, Brian A.; Dudley, Joel T.

    2016-05-01

    Secondary use of electronic health records (EHRs) promises to advance clinical research and better inform clinical decision making. Challenges in summarizing and representing patient data prevent widespread practice of predictive modeling using EHRs. Here we present a novel unsupervised deep feature learning method to derive a general-purpose patient representation from EHR data that facilitates clinical predictive modeling. In particular, a three-layer stack of denoising autoencoders was used to capture hierarchical regularities and dependencies in the aggregated EHRs of about 700,000 patients from the Mount Sinai data warehouse. The result is a representation we name “deep patient”. We evaluated this representation as broadly predictive of health states by assessing the probability of patients to develop various diseases. We performed evaluation using 76,214 test patients comprising 78 diseases from diverse clinical domains and temporal windows. Our results significantly outperformed those achieved using representations based on raw EHR data and alternative feature learning strategies. Prediction performance for severe diabetes, schizophrenia, and various cancers were among the top performing. These findings indicate that deep learning applied to EHRs can derive patient representations that offer improved clinical predictions, and could provide a machine learning framework for augmenting clinical decision systems.

  3. Voxel-based clustered imaging by multiparameter diffusion tensor images for glioma grading.

    Science.gov (United States)

    Inano, Rika; Oishi, Naoya; Kunieda, Takeharu; Arakawa, Yoshiki; Yamao, Yukihiro; Shibata, Sumiya; Kikuchi, Takayuki; Fukuyama, Hidenao; Miyamoto, Susumu

    2014-01-01

    Gliomas are the most common intra-axial primary brain tumour; therefore, predicting glioma grade would influence therapeutic strategies. Although several methods based on single or multiple parameters from diagnostic images exist, a definitive method for pre-operatively determining glioma grade remains unknown. We aimed to develop an unsupervised method using multiple parameters from pre-operative diffusion tensor images for obtaining a clustered image that could enable visual grading of gliomas. Fourteen patients with low-grade gliomas and 19 with high-grade gliomas underwent diffusion tensor imaging and three-dimensional T1-weighted magnetic resonance imaging before tumour resection. Seven features including diffusion-weighted imaging, fractional anisotropy, first eigenvalue, second eigenvalue, third eigenvalue, mean diffusivity and raw T2 signal with no diffusion weighting, were extracted as multiple parameters from diffusion tensor imaging. We developed a two-level clustering approach for a self-organizing map followed by the K-means algorithm to enable unsupervised clustering of a large number of input vectors with the seven features for the whole brain. The vectors were grouped by the self-organizing map as protoclusters, which were classified into the smaller number of clusters by K-means to make a voxel-based diffusion tensor-based clustered image. Furthermore, we also determined if the diffusion tensor-based clustered image was really helpful for predicting pre-operative glioma grade in a supervised manner. The ratio of each class in the diffusion tensor-based clustered images was calculated from the regions of interest manually traced on the diffusion tensor imaging space, and the common logarithmic ratio scales were calculated. We then applied support vector machine as a classifier for distinguishing between low- and high-grade gliomas. Consequently, the sensitivity, specificity, accuracy and area under the curve of receiver operating characteristic

  4. Cluster analysis of polymers using laser-induced breakdown spectroscopy with K-means

    Science.gov (United States)

    Yangmin, GUO; Yun, TANG; Yu, DU; Shisong, TANG; Lianbo, GUO; Xiangyou, LI; Yongfeng, LU; Xiaoyan, ZENG

    2018-06-01

    Laser-induced breakdown spectroscopy (LIBS) combined with K-means algorithm was employed to automatically differentiate industrial polymers under atmospheric conditions. The unsupervised learning algorithm K-means were utilized for the clustering of LIBS dataset measured from twenty kinds of industrial polymers. To prevent the interference from metallic elements, three atomic emission lines (C I 247.86 nm , H I 656.3 nm, and O I 777.3 nm) and one molecular line C–N (0, 0) 388.3 nm were used. The cluster analysis results were obtained through an iterative process. The Davies–Bouldin index was employed to determine the initial number of clusters. The average relative standard deviation values of characteristic spectral lines were used as the iterative criterion. With the proposed approach, the classification accuracy for twenty kinds of industrial polymers achieved 99.6%. The results demonstrated that this approach has great potential for industrial polymers recycling by LIBS.

  5. Classification of high-resolution multi-swath hyperspectral data using Landsat 8 surface reflectance data as a calibration target and a novel histogram based unsupervised classification technique to determine natural classes from biophysically relevant fit parameters

    Science.gov (United States)

    McCann, C.; Repasky, K. S.; Morin, M.; Lawrence, R. L.; Powell, S. L.

    2016-12-01

    Compact, cost-effective, flight-based hyperspectral imaging systems can provide scientifically relevant data over large areas for a variety of applications such as ecosystem studies, precision agriculture, and land management. To fully realize this capability, unsupervised classification techniques based on radiometrically-calibrated data that cluster based on biophysical similarity rather than simply spectral similarity are needed. An automated technique to produce high-resolution, large-area, radiometrically-calibrated hyperspectral data sets based on the Landsat surface reflectance data product as a calibration target was developed and applied to three subsequent years of data covering approximately 1850 hectares. The radiometrically-calibrated data allows inter-comparison of the temporal series. Advantages of the radiometric calibration technique include the need for minimal site access, no ancillary instrumentation, and automated processing. Fitting the reflectance spectra of each pixel using a set of biophysically relevant basis functions reduces the data from 80 spectral bands to 9 parameters providing noise reduction and data compression. Examination of histograms of these parameters allows for determination of natural splitting into biophysical similar clusters. This method creates clusters that are similar in terms of biophysical parameters, not simply spectral proximity. Furthermore, this method can be applied to other data sets, such as urban scenes, by developing other physically meaningful basis functions. The ability to use hyperspectral imaging for a variety of important applications requires the development of data processing techniques that can be automated. The radiometric-calibration combined with the histogram based unsupervised classification technique presented here provide one potential avenue for managing big-data associated with hyperspectral imaging.

  6. Dynamic Hierarchical Sleep Scheduling for Wireless Ad-Hoc Sensor Networks

    OpenAIRE

    Chih-Yu Wen; Ying-Chih Chen

    2009-01-01

    This paper presents two scheduling management schemes for wireless sensor networks, which manage the sensors by utilizing the hierarchical network structure and allocate network resources efficiently. A local criterion is used to simultaneously establish the sensing coverage and connectivity such that dynamic cluster-based sleep scheduling can be achieved. The proposed schemes are simulated and analyzed to abstract the network behaviors in a number of settings. The experimental results show t...

  7. AHIMSA - Ad hoc histogram information measure sensing algorithm for feature selection in the context of histogram inspired clustering techniques

    Science.gov (United States)

    Dasarathy, B. V.

    1976-01-01

    An algorithm is proposed for dimensionality reduction in the context of clustering techniques based on histogram analysis. The approach is based on an evaluation of the hills and valleys in the unidimensional histograms along the different features and provides an economical means of assessing the significance of the features in a nonparametric unsupervised data environment. The method has relevance to remote sensing applications.

  8. Formal And Informal Macro-Regional Transport Clusters As A Primary Step In The Design And Implementation Of Cluster-Based Strategies

    Directory of Open Access Journals (Sweden)

    Nežerenko Olga

    2015-09-01

    Full Text Available The aim of the study is the identification of a formal macro-regional transport and logistics cluster and its development trends on a macro-regional level in 2007-2011 by means of the hierarchical cluster analysis. The central approach of the study is based on two concepts: 1 the concept of formal and informal macro-regions, and 2 the concept of clustering which is based on the similarities shared by the countries of a macro-region and tightly related to the concept of macro-region. The authors seek to answer the question whether the formation of a formal transport cluster could provide the BSR a stable competitive position in the global transportation and logistics market.

  9. Evolution of the cluster X-ray luminosity function

    DEFF Research Database (Denmark)

    Mullis, C.R.; Vikhlinin, A.; Henry, J.P.

    2004-01-01

    We report measurements of the cluster X-ray luminosity function out to z = 0.8 based on the final sample of 201 galaxy systems from the 160 Square Degree ROSAT Cluster Survey. There is little evidence for any measurable change in cluster abundance out to z similar to 0.6 at luminosities of less...... than a few times 10(44) h(50)(-2) ergs s(-1) (0.5 - 2.0 keV). However, for 0.6 cluster deficit using integrated number counts...... independently confirm the presence of evolution. Whereas the bulk of the cluster population does not evolve, the most luminous and presumably most massive structures evolve appreciably between z = 0.8 and the present. Interpreted in the context of hierarchical structure formation, we are probing sufficiently...

  10. AIDEN: A Density Conscious Artificial Immune System for Automatic Discovery of Arbitrary Shape Clusters in Spatial Patterns

    Directory of Open Access Journals (Sweden)

    Vishwambhar Pathak

    2012-11-01

    Full Text Available Recent efforts in modeling of dynamics of the natural immune cells leading to artificial immune systems (AIS have ignited contemporary research interest in finding out its analogies to real world problems. The AIS models have been vastly exploited to develop dependable robust
    solutions to clustering. Most of the traditional clustering methods bear limitations in their capability to detect clusters of arbitrary shapes in a fully unsupervised manner. In this paper the recognition and communication dynamics of T Cell Receptors, the recognizing elements in innate immune
    system, has been modeled with a kernel density estimation method. The model has been shown to successfully discover non spherical clusters in spatial patterns. Modeling the cohesion of the antibodies and pathogens with ‘local influence’ measure inducts comprehensive extension of the
    antibody representation ball (ARB, which in turn corresponds to controlled expansion of clusters and prevents overfitting.

  11. Unsupervised Learning of Spatiotemporal Features by Video Completion

    OpenAIRE

    Nallabolu, Adithya Reddy

    2017-01-01

    In this work, we present an unsupervised representation learning approach for learning rich spatiotemporal features from videos without the supervision from semantic labels. We propose to learn the spatiotemporal features by training a 3D convolutional neural network (CNN) using video completion as a surrogate task. Using a large collection of unlabeled videos, we train the CNN to predict the missing pixels of a spatiotemporal hole given the remaining parts of the video through minimizing per...

  12. HIERARCHICAL FRAGMENTATION AND JET-LIKE OUTFLOWS IN IRDC G28.34+0.06: A GROWING MASSIVE PROTOSTAR CLUSTER

    International Nuclear Information System (INIS)

    Wang Ke; Wu Yuefang; Zhang Huawei; Zhang Qizhou

    2011-01-01

    We present Submillimeter Array (SMA) λ = 0.88 mm observations of an infrared dark cloud G28.34+0.06. Located in the quiescent southern part of the G28.34 cloud, the region of interest is a massive (>10 3 M sun ) molecular clump P1 with a luminosity of ∼10 3 L sun , where our previous SMA observations at 1.3 mm have revealed a string of five dust cores of 22-64 M sun along the 1 pc IR-dark filament. The cores are well aligned at a position angle (P.A.) of 48 deg. and regularly spaced at an average projected separation of 0.16 pc. The new high-resolution, high-sensitivity 0.88 mm image further resolves the five cores into 10 compact condensations of 1.4-10.6 M sun , with sizes of a few thousand AU. The spatial structure at clump (∼1 pc) and core (∼0.1 pc) scales indicates a hierarchical fragmentation. While the clump fragmentation is consistent with a cylindrical collapse, the observed fragment masses are much larger than the expected thermal Jeans masses. All the cores are driving CO (3-2) outflows up to 38 km s -1 , the majority of which are bipolar, jet-like outflows. The moderate luminosity of the P1 clump sets a limit on the mass of protostars of 3-7 M sun . Because of the large reservoir of dense molecular gas in the immediate medium and ongoing accretion as evident by the jet-like outflows, we speculate that P1 will grow and eventually form a massive star cluster. This study provides a first glimpse of massive, clustered star formation that currently undergoes through an intermediate-mass stage.

  13. Segmentation methodology for automated classification and differentiation of soft tissues in multiband images of high-resolution ultrasonic transmission tomography.

    Science.gov (United States)

    Jeong, Jeong-Won; Shin, Dae C; Do, Synho; Marmarelis, Vasilis Z

    2006-08-01

    This paper presents a novel segmentation methodology for automated classification and differentiation of soft tissues using multiband data obtained with the newly developed system of high-resolution ultrasonic transmission tomography (HUTT) for imaging biological organs. This methodology extends and combines two existing approaches: the L-level set active contour (AC) segmentation approach and the agglomerative hierarchical kappa-means approach for unsupervised clustering (UC). To prevent the trapping of the current iterative minimization AC algorithm in a local minimum, we introduce a multiresolution approach that applies the level set functions at successively increasing resolutions of the image data. The resulting AC clusters are subsequently rearranged by the UC algorithm that seeks the optimal set of clusters yielding the minimum within-cluster distances in the feature space. The presented results from Monte Carlo simulations and experimental animal-tissue data demonstrate that the proposed methodology outperforms other existing methods without depending on heuristic parameters and provides a reliable means for soft tissue differentiation in HUTT images.

  14. Hierarchical drivers of reef-fish metacommunity structure.

    Science.gov (United States)

    MacNeil, M Aaron; Graham, Nicholas A J; Polunin, Nicholas V C; Kulbicki, Michel; Galzin, René; Harmelin-Vivien, Mireille; Rushton, Steven P

    2009-01-01

    multiple spatial scales; and (3) inter-atoll connectedness was poorly correlated with the nonrandom clustering of reef-fish species. These results demonstrate the importance of modeling hierarchical data and processes in understanding reef-fish metacommunity structure.

  15. Supervised and unsupervised condition monitoring of non-stationary acoustic emission signals

    DEFF Research Database (Denmark)

    Sigurdsson, Sigurdur; Pontoppidan, Niels Henrik; Larsen, Jan

    2005-01-01

    condition changes across load changes. In this paper we approach this load interpolation problem with supervised and unsupervised learning, i.e. model with normal and fault examples and normal examples only, respectively. We apply non-linear methods for the learning of engine condition changes. Both...

  16. Evaluating unsupervised methods to size and classify suspended particles using digital in-line holography

    Science.gov (United States)

    Davies, Emlyn J.; Buscombe, Daniel D.; Graham, George W.; Nimmo-Smith, W. Alex M.

    2015-01-01

    Substantial information can be gained from digital in-line holography of marine particles, eliminating depth-of-field and focusing errors associated with standard lens-based imaging methods. However, for the technique to reach its full potential in oceanographic research, fully unsupervised (automated) methods are required for focusing, segmentation, sizing and classification of particles. These computational challenges are the subject of this paper, in which we draw upon data collected using a variety of holographic systems developed at Plymouth University, UK, from a significant range of particle types, sizes and shapes. A new method for noise reduction in reconstructed planes is found to be successful in aiding particle segmentation and sizing. The performance of an automated routine for deriving particle characteristics (and subsequent size distributions) is evaluated against equivalent size metrics obtained by a trained operative measuring grain axes on screen. The unsupervised method is found to be reliable, despite some errors resulting from over-segmentation of particles. A simple unsupervised particle classification system is developed, and is capable of successfully differentiating sand grains, bubbles and diatoms from within the surf-zone. Avoiding miscounting bubbles and biological particles as sand grains enables more accurate estimates of sand concentrations, and is especially important in deployments of particle monitoring instrumentation in aerated water. Perhaps the greatest potential for further development in the computational aspects of particle holography is in the area of unsupervised particle classification. The simple method proposed here provides a foundation upon which further development could lead to reliable identification of more complex particle populations, such as those containing phytoplankton, zooplankton, flocculated cohesive sediments and oil droplets.

  17. Unsupervised categorization with individuals diagnosed as having moderate traumatic brain injury: Over-selective responding.

    Science.gov (United States)

    Edwards, Darren J; Wood, Rodger

    2016-01-01

    This study explored over-selectivity (executive dysfunction) using a standard unsupervised categorization task. Over-selectivity has been demonstrated using supervised categorization procedures (where training is given); however, little has been done in the way of unsupervised categorization (without training). A standard unsupervised categorization task was used to assess levels of over-selectivity in a traumatic brain injury (TBI) population. Individuals with TBI were selected from the Tertiary Traumatic Brain Injury Clinic at Swansea University and were asked to categorize two-dimensional items (pictures on cards), into groups that they felt were most intuitive, and without any learning (feedback from experimenter). This was compared against categories made by a control group for the same task. The findings of this study demonstrate that individuals with TBI had deficits for both easy and difficult categorization sets, as indicated by a larger amount of one-dimensional sorting compared to control participants. Deficits were significantly greater for the easy condition. The implications of these findings are discussed in the context of over-selectivity, and the processes that underlie this deficit. Also, the implications for using this procedure as a screening measure for over-selectivity in TBI are discussed.

  18. tclust: An R Package for a Trimming Approach to Cluster Analysis

    Directory of Open Access Journals (Sweden)

    2012-04-01

    Full Text Available Outlying data can heavily influence standard clustering methods. At the same time, clustering principles can be useful when robustifying statistical procedures. These two reasons motivate the development of feasible robust model-based clustering approaches. With this in mind, an R package for performing non-hierarchical robust clustering, called tclust, is presented here. Instead of trying to “fit” noisy data, a proportion α of the most outlying observations is trimmed. The tclust package efficiently handles different cluster scatter constraints. Graphical exploratory tools are also provided to help the user make sensible choices for the trimming proportion as well as the number of clusters to search for.

  19. Six weeks of unsupervised Nintendo Wii Fit gaming is effective at improving balance in independent older adults.

    Science.gov (United States)

    Nicholson, Vaughan Patrick; McKean, Mark; Lowe, John; Fawcett, Christine; Burkett, Brendan

    2015-01-01

    To determine the effectiveness of unsupervised Nintendo Wii Fit balance training in older adults. Forty-one older adults were recruited from local retirement villages and educational settings to participate in a six-week two-group repeated measures study. The Wii group (n = 19, 75 ± 6 years) undertook 30 min of unsupervised Wii balance gaming three times per week in their retirement village while the comparison group (n = 22, 74 ± 5 years) continued with their usual exercise program. Participants' balance abilities were assessed pre- and postintervention. The Wii Fit group demonstrated significant improvements (P balance, lateral reach (left and right), and gait speed compared with the comparison group. Reported levels of enjoyment following game play increased during the study. Six weeks of unsupervised Wii balance training is an effective modality for improving balance in independent older adults.

  20. Function approximation using combined unsupervised and supervised learning.

    Science.gov (United States)

    Andras, Peter

    2014-03-01

    Function approximation is one of the core tasks that are solved using neural networks in the context of many engineering problems. However, good approximation results need good sampling of the data space, which usually requires exponentially increasing volume of data as the dimensionality of the data increases. At the same time, often the high-dimensional data is arranged around a much lower dimensional manifold. Here we propose the breaking of the function approximation task for high-dimensional data into two steps: (1) the mapping of the high-dimensional data onto a lower dimensional space corresponding to the manifold on which the data resides and (2) the approximation of the function using the mapped lower dimensional data. We use over-complete self-organizing maps (SOMs) for the mapping through unsupervised learning, and single hidden layer neural networks for the function approximation through supervised learning. We also extend the two-step procedure by considering support vector machines and Bayesian SOMs for the determination of the best parameters for the nonlinear neurons in the hidden layer of the neural networks used for the function approximation. We compare the approximation performance of the proposed neural networks using a set of functions and show that indeed the neural networks using combined unsupervised and supervised learning outperform in most cases the neural networks that learn the function approximation using the original high-dimensional data.

  1. A roadmap of clustering algorithms: finding a match for a biomedical application.

    Science.gov (United States)

    Andreopoulos, Bill; An, Aijun; Wang, Xiaogang; Schroeder, Michael

    2009-05-01

    Clustering is ubiquitously applied in bioinformatics with hierarchical clustering and k-means partitioning being the most popular methods. Numerous improvements of these two clustering methods have been introduced, as well as completely different approaches such as grid-based, density-based and model-based clustering. For improved bioinformatics analysis of data, it is important to match clusterings to the requirements of a biomedical application. In this article, we present a set of desirable clustering features that are used as evaluation criteria for clustering algorithms. We review 40 different clustering algorithms of all approaches and datatypes. We compare algorithms on the basis of desirable clustering features, and outline algorithms' benefits and drawbacks as a basis for matching them to biomedical applications.

  2. Statistical discovery of site inter-dependencies in sub-molecular hierarchical protein structuring.

    Science.gov (United States)

    Durston, Kirk K; Chiu, David Ky; Wong, Andrew Kc; Li, Gary Cl

    2012-07-13

    Much progress has been made in understanding the 3D structure of proteins using methods such as NMR and X-ray crystallography. The resulting 3D structures are extremely informative, but do not always reveal which sites and residues within the structure are of special importance. Recently, there are indications that multiple-residue, sub-domain structural relationships within the larger 3D consensus structure of a protein can be inferred from the analysis of the multiple sequence alignment data of a protein family. These intra-dependent clusters of associated sites are used to indicate hierarchical inter-residue relationships within the 3D structure. To reveal the patterns of associations among individual amino acids or sub-domain components within the structure, we apply a k-modes attribute (aligned site) clustering algorithm to the ubiquitin and transthyretin families in order to discover associations among groups of sites within the multiple sequence alignment. We then observe what these associations imply within the 3D structure of these two protein families. The k-modes site clustering algorithm we developed maximizes the intra-group interdependencies based on a normalized mutual information measure. The clusters formed correspond to sub-structural components or binding and interface locations. Applying this data-directed method to the ubiquitin and transthyretin protein family multiple sequence alignments as a test bed, we located numerous interesting associations of interdependent sites. These clusters were then arranged into cluster tree diagrams which revealed four structural sub-domains within the single domain structure of ubiquitin and a single large sub-domain within transthyretin associated with the interface among transthyretin monomers. In addition, several clusters of mutually interdependent sites were discovered for each protein family, each of which appear to play an important role in the molecular structure and/or function. Our results

  3. Accurate detection of hierarchical communities in complex networks based on nonlinear dynamical evolution

    Science.gov (United States)

    Zhuo, Zhao; Cai, Shi-Min; Tang, Ming; Lai, Ying-Cheng

    2018-04-01

    One of the most challenging problems in network science is to accurately detect communities at distinct hierarchical scales. Most existing methods are based on structural analysis and manipulation, which are NP-hard. We articulate an alternative, dynamical evolution-based approach to the problem. The basic principle is to computationally implement a nonlinear dynamical process on all nodes in the network with a general coupling scheme, creating a networked dynamical system. Under a proper system setting and with an adjustable control parameter, the community structure of the network would "come out" or emerge naturally from the dynamical evolution of the system. As the control parameter is systematically varied, the community hierarchies at different scales can be revealed. As a concrete example of this general principle, we exploit clustered synchronization as a dynamical mechanism through which the hierarchical community structure can be uncovered. In particular, for quite arbitrary choices of the nonlinear nodal dynamics and coupling scheme, decreasing the coupling parameter from the global synchronization regime, in which the dynamical states of all nodes are perfectly synchronized, can lead to a weaker type of synchronization organized as clusters. We demonstrate the existence of optimal choices of the coupling parameter for which the synchronization clusters encode accurate information about the hierarchical community structure of the network. We test and validate our method using a standard class of benchmark modular networks with two distinct hierarchies of communities and a number of empirical networks arising from the real world. Our method is computationally extremely efficient, eliminating completely the NP-hard difficulty associated with previous methods. The basic principle of exploiting dynamical evolution to uncover hidden community organizations at different scales represents a "game-change" type of approach to addressing the problem of community

  4. Consensus clustering approach to group brain connectivity matrices

    Directory of Open Access Journals (Sweden)

    Javier Rasero

    2017-10-01

    Full Text Available A novel approach rooted on the notion of consensus clustering, a strategy developed for community detection in complex networks, is proposed to cope with the heterogeneity that characterizes connectivity matrices in health and disease. The method can be summarized as follows: (a define, for each node, a distance matrix for the set of subjects by comparing the connectivity pattern of that node in all pairs of subjects; (b cluster the distance matrix for each node; (c build the consensus network from the corresponding partitions; and (d extract groups of subjects by finding the communities of the consensus network thus obtained. Different from the previous implementations of consensus clustering, we thus propose to use the consensus strategy to combine the information arising from the connectivity patterns of each node. The proposed approach may be seen either as an exploratory technique or as an unsupervised pretraining step to help the subsequent construction of a supervised classifier. Applications on a toy model and two real datasets show the effectiveness of the proposed methodology, which represents heterogeneity of a set of subjects in terms of a weighted network, the consensus matrix.

  5. Unsupervised Word Mapping Using Structural Similarities in Monolingual Embeddings

    OpenAIRE

    Aldarmaki, Hanan; Mohan, Mahesh; Diab, Mona

    2017-01-01

    Most existing methods for automatic bilingual dictionary induction rely on prior alignments between the source and target languages, such as parallel corpora or seed dictionaries. For many language pairs, such supervised alignments are not readily available. We propose an unsupervised approach for learning a bilingual dictionary for a pair of languages given their independently-learned monolingual word embeddings. The proposed method exploits local and global structures in monolingual vector ...

  6. Landsat TM band 431 combine on clustering analysis for pattern recognition land use using idrisi 4.2 software

    International Nuclear Information System (INIS)

    Wiweka, Arief H.; Izzawati, Tjahyaningsih A.

    1997-01-01

    The recognition of earth object's pattern which is recorded on remote sensing digital image can do by classification process based on the group of spectral pixel value. The spectral assessment on a spatial which represent the object characteristic can be helped through supervised or unsupervised. On certain case, there no media, such as maps, airborne, photo, the capability of field observation and the knowledge of object's location. Classification process can be done by clustering. The group of pixel based on the wide of the whole value interval of spectral image, then the class group base on the desired accuracy. The clustering method in Idris 4.2 software equipments are sequential method, statistic, iso data, and RGB. The clustering existence can help pre-process pattern recognition

  7. Radiation-associated breast tumors display a distinct gene expression profile

    DEFF Research Database (Denmark)

    Broeks, Annegien; Braaf, Linde M; Wessels, Lodewyk F A

    2010-01-01

    PURPOSE: Women who received irradiation for Hodgkin's lymphoma have a strong increased risk for developing breast cancer. Approximately 90% of the breast cancers in these patients can be attributed to their radiation treatment, rendering such series extremely useful to determine whether a common...... radiation-associated cause underlies the carcinogenic process. METHODS AND MATERIALS: In this study we used gene expression profiling technology to assess gene expression changes in radiation-associated breast tumors compared with a set of control breast tumors of women unexposed to radiation, diagnosed...... at the same age. RNA was obtained from fresh frozen tissue samples from 22 patients who developed breast cancer after Hodgkin's lymphoma (BfHL) and from 20 control breast tumors. RESULTS: Unsupervised hierarchical clustering of the profile data resulted in a clustering of the radiation-associated tumors...

  8. Typing of unknown microorganisms based on quantitative analysis of fatty acids by mass spectrometry and hierarchical clustering

    Energy Technology Data Exchange (ETDEWEB)

    Li Tingting; Dai Ling; Li Lun; Hu Xuejiao; Dong Linjie; Li Jianjian; Salim, Sule Khalfan; Fu Jieying [Key Laboratory of Pesticides and Chemical Biology, Ministry of Education, College of Chemistry, Central China Normal University, Wuhan, Hubei 430079 (China); Zhong Hongying, E-mail: hyzhong@mail.ccnu.edu.cn [Key Laboratory of Pesticides and Chemical Biology, Ministry of Education, College of Chemistry, Central China Normal University, Wuhan, Hubei 430079 (China)

    2011-01-17

    Rapid identification of unknown microorganisms of clinical and agricultural importance is not only critical for accurate diagnosis of infections but also essential for appropriate and prompt treatment. We describe here a rapid method for microorganisms typing based on quantitative analysis of fatty acids by iFAT approach (Isotope-coded Fatty Acid Transmethylation). In this work, lyophilized cell lysates were directly mixed with 0.5 M NaOH solution in d3-methanol and n-hexane. After 1 min of ultrasonication, the top n-hexane layer was combined with a mixture of standard d0-methanol derived fatty acid methylesters with known concentration. Measurement of intensity ratios of d3/d0 labeled fragment ion and molecular ion pairs at the corresponding target fatty acids provides a quantitative basis for hierarchical clustering. In the resultant dendrogram, the Euclidean distance between unknown species and known species quantitatively reveals their differences or shared similarities in fatty acid related pathways. It is of particular interest to apply this method for typing fungal species because fungi has distinguished lipid biosynthetic pathways that have been targeted for lots of drugs or fungicides compared with bacteria and animals. The proposed method has no dependence on the availability of genome or proteome databases. Therefore, it is can be applicable for a broad range of unknown microorganisms or mutant species.

  9. Clusters and Groups of Galaxies : International Meeting

    CERN Document Server

    Giuricin, G; Mezzetti, M

    1984-01-01

    The large-scale structure of the Universe and systems Clusters, and Groups of galaxies are topics like Superclusters, They fully justify the meeting on "Clusters of great interest. and Groups of Galaxies". The topics covered included the spatial distribution and the clustering of galaxies; the properties of Superclusters, Clusters and Groups of galaxies; radio and X-ray observations; the problem of unseen matter; theories concerning hierarchical clustering, pancakes, cluster and galaxy formation and evolution. The meeting was held at the International Center for Theoretical Physics in Trieste (Italy) from September 13 to September 16, 1983. It was attended by about 150 participants from 22 nations who presented 67 invited lectures (il) and contributed papers (cp), and 45 poster papers (pp). The Scientific Organizing Committee consisted of F. Bertola, P. Biermann, A. Cavaliere, N. Dallaporta, D. Gerba1, M. Hack, J . V . Peach, D. Sciama (Chairman), G. Setti, M. Tarenghi. We are particularly indebted to D. Scia...

  10. Validation of a free software for unsupervised assessment of abdominal fat in MRI.

    Science.gov (United States)

    Maddalo, Michele; Zorza, Ivan; Zubani, Stefano; Nocivelli, Giorgio; Calandra, Giulio; Soldini, Pierantonio; Mascaro, Lorella; Maroldi, Roberto

    2017-05-01

    To demonstrate the accuracy of an unsupervised (fully automated) software for fat segmentation in magnetic resonance imaging. The proposed software is a freeware solution developed in ImageJ that enables the quantification of metabolically different adipose tissues in large cohort studies. The lumbar part of the abdomen (19cm in craniocaudal direction, centered in L3) of eleven healthy volunteers (age range: 21-46years, BMI range: 21.7-31.6kg/m 2 ) was examined in a breath hold on expiration with a GE T1 Dixon sequence. Single-slice and volumetric data were considered for each subject. The results of the visceral and subcutaneous adipose tissue assessments obtained by the unsupervised software were compared to supervised segmentations of reference. The associated statistical analysis included Pearson correlations, Bland-Altman plots and volumetric differences (VD % ). Values calculated by the unsupervised software significantly correlated with corresponding supervised segmentations of reference for both subcutaneous adipose tissue - SAT (R=0.9996, psoftware is capable of segmenting the metabolically different adipose tissues with a high degree of accuracy. This free add-on software for ImageJ can easily have a widespread and enable large-scale population studies regarding the adipose tissue and its related diseases. Copyright © 2017 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.

  11. Recommending the heterogeneous cluster type multi-processor system computing

    International Nuclear Information System (INIS)

    Iijima, Nobukazu

    2010-01-01

    Real-time reactor simulator had been developed by reusing the equipment of the Musashi reactor and its performance improvement became indispensable for research tools to increase sampling rate with introduction of arithmetic units using multi-Digital Signal Processor(DSP) system (cluster). In order to realize the heterogeneous cluster type multi-processor system computing, combination of two kinds of Control Processor (CP) s, Cluster Control Processor (CCP) and System Control Processor (SCP), were proposed with Large System Control Processor (LSCP) for hierarchical cluster if needed. Faster computing performance of this system was well evaluated by simulation results for simultaneous execution of plural jobs and also pipeline processing between clusters, which showed the system led to effective use of existing system and enhancement of the cost performance. (T. Tanaka)

  12. Simultaneous determination of 19 flavonoids in commercial trollflowers by using high-performance liquid chromatography and classification of samples by hierarchical clustering analysis.

    Science.gov (United States)

    Song, Zhiling; Hashi, Yuki; Sun, Hongyang; Liang, Yi; Lan, Yuexiang; Wang, Hong; Chen, Shizhong

    2013-12-01

    The flowers of Trollius species, named Jin Lianhua in Chinese, are widely used traditional Chinese herbs with vital biological activity that has been used for several decades in China to treat upper respiratory infections, pharyngitis, tonsillitis, and bronchitis. We developed a rapid and reliable method for simultaneous quantitative analysis of 19 flavonoids in trollflowers by using high-performance liquid chromatography (HPLC). Chromatography was performed on Inertsil ODS-3 C18 column, with gradient elution methanol-acetonitrile-water with 0.02% (v/v) formic acid. Content determination was used to evaluate the quality of commercial trollflowers from different regions in China, while three Trollius species (Trollius chinensis Bunge, Trollius ledebouri Reichb, Trollius buddae Schipcz) were explicitly distinguished by using hierarchical clustering analysis. The linearity, precision, accuracy, limit of detection, and limit of quantification were validated for the quantification method, which proved sensitive, accurate and reproducible indicating that the proposed approach was applicable for the routine analysis and quality control of trollflowers. © 2013.

  13. Superhydrophobic polytetrafluoroethylene thin films with hierarchical roughness deposited using a single step vapor phase technique

    International Nuclear Information System (INIS)

    Gupta, Sushant; Arjunan, Arul Chakkaravarthi; Deshpande, Sameer; Seal, Sudipta; Singh, Deepika; Singh, Rajiv K.

    2009-01-01

    Superhydrophobic polytetrafluoroethylene films with hierarchical surface roughness were deposited using pulse electron deposition technique. We were able to modulate roughness of the deposited films by controlling the beam energy and hence the electron penetration depth. The films deposited at higher beam energy showed contact angle as high as 166 o . The scanning electron and atomic force microscope studies revealed clustered growth and two level sub-micron asperities on films deposited at higher energies. Such dual-scale hierarchical roughness and heterogeneities at the water-surface interface was attributed to the observed contact angle and thus its superhydrophobic nature.

  14. Superhydrophobic polytetrafluoroethylene thin films with hierarchical roughness deposited using a single step vapor phase technique

    Energy Technology Data Exchange (ETDEWEB)

    Gupta, Sushant, E-mail: sushant3@ufl.ed [Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611 (United States); Arjunan, Arul Chakkaravarthi [Sinmat Incorporated, 2153 SE Hawthorne Road, 129, Gainesville, Florida 32641 (United States); Deshpande, Sameer; Seal, Sudipta [Advanced Material Processing and Analysis Center, University of Central Florida, Orlando, Florida 32816 (United States); Singh, Deepika [Sinmat Incorporated, 2153 SE Hawthorne Road, 129, Gainesville, Florida 32641 (United States); Singh, Rajiv K. [Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611 (United States)

    2009-06-30

    Superhydrophobic polytetrafluoroethylene films with hierarchical surface roughness were deposited using pulse electron deposition technique. We were able to modulate roughness of the deposited films by controlling the beam energy and hence the electron penetration depth. The films deposited at higher beam energy showed contact angle as high as 166{sup o}. The scanning electron and atomic force microscope studies revealed clustered growth and two level sub-micron asperities on films deposited at higher energies. Such dual-scale hierarchical roughness and heterogeneities at the water-surface interface was attributed to the observed contact angle and thus its superhydrophobic nature.

  15. Investigation of major international and Turkish companies via hierarchical methods and bootstrap approach

    Science.gov (United States)

    Kantar, E.; Deviren, B.; Keskin, M.

    2011-11-01

    We present a study, within the scope of econophysics, of the hierarchical structure of 98 among the largest international companies including 18 among the largest Turkish companies, namely Banks, Automobile, Software-hardware, Telecommunication Services, Energy and the Oil-Gas sectors, viewed as a network of interacting companies. We analyze the daily time series data of the Boerse-Frankfurt and Istanbul Stock Exchange. We examine the topological properties among the companies over the period 2006-2010 by using the concept of hierarchical structure methods (the minimal spanning tree (MST) and the hierarchical tree (HT)). The period is divided into three subperiods, namely 2006-2007, 2008 which was the year of global economic crisis, and 2009-2010, in order to test various time-windows and observe temporal evolution. We carry out bootstrap analyses to associate the value of statistical reliability to the links of the MSTs and HTs. We also use average linkage clustering analysis (ALCA) in order to better observe the cluster structure. From these studies, we find that the interactions among the Banks/Energy sectors and the other sectors were reduced after the global economic crisis; hence the effects of the Banks and Energy sectors on the correlations of all companies were decreased. Telecommunication Services were also greatly affected by the crisis. We also observed that the Automobile and Banks sectors, including Turkish companies as well as some companies from the USA, Japan and Germany were strongly correlated with each other in all periods.

  16. Approximate fuzzy C-means (AFCM) cluster analysis of medical magnetic resonance image (MRI) data

    International Nuclear Information System (INIS)

    DelaPaz, R.L.; Chang, P.J.; Bernstein, R.; Dave, J.V.

    1987-01-01

    The authors describe the application of an approximate fuzzy C-means (AFCM) clustering algorithm as a data dimension reduction approach to medical magnetic resonance images (MRI). Image data consisted of one T1-weighted, two T2-weighted, and one T2*-weighted (magnetic susceptibility) image for each cranial study and a matrix of 10 images generated from 10 combinations of TE and TR for each body lymphoma study. All images were obtained with a 1.5 Tesla imaging system (GE Signa). Analyses were performed on over 100 MR image sets with a variety of pathologies. The cluster analysis was operated in an unsupervised mode and computational overhead was minimized by utilizing a table look-up approach without adversely affecting accuracy. Image data were first segmented into 2 coarse clusters, each of which was then subdivided into 16 fine clusters. The final tissue classifications were presented as color-coded anatomically-mapped images and as two and three dimensional displays of cluster center data in selected feature space (minimum spanning tree). Fuzzy cluster analysis appears to be a clinically useful dimension reduction technique which results in improved diagnostic specificity of medical magnetic resonance images

  17. Enhanced HMAX model with feedforward feature learning for multiclass categorization

    Directory of Open Access Journals (Sweden)

    Yinlin eLi

    2015-10-01

    Full Text Available In recent years, the interdisciplinary research between neuroscience and computer vision has promoted the development in both fields. Many biologically inspired visual models are proposed, and among them, the Hierarchical Max-pooling model (HMAX is a feedforward model mimicking the structures and functions of V1 to posterior inferotemporal (PIT layer of the primate visual cortex, which could generate a series of position- and scale- invariant features. However, it could be improved with attention modulation and memory processing, which are two important properties of the primate visual cortex. Thus, in this paper, based on recent biological research on the primate visual cortex, we still mimic the first 100-150 milliseconds of visual cognition to enhance the HMAX model, which mainly focuses on the unsupervised feedforward feature learning process. The main modifications are as follows: 1 To mimic the attention modulation mechanism of V1 layer, a bottom-up saliency map is computed in the S1 layer of the HMAX model, which can support the initial feature extraction for memory processing; 2 To mimic the learning, clustering and short-term memory to long-term memory conversion abilities of V2 and IT, an unsupervised iterative clustering method is used to learn clusters with multiscale middle level patches, which are taken as long-term memory; 3 Inspired by the multiple feature encoding mode of the primate visual cortex, information including color, orientation, and spatial position are encoded in different layers of the HMAX model progressively. By adding a softmax layer at the top of the model, multiclass categorization experiments can be conducted, and the results on Caltech101 show that the enhanced model with a smaller memory size exhibits higher accuracy than the original HMAX model, and could also achieve better accuracy than other unsupervised feature learning methods in multiclass categorization task.

  18. Enhanced HMAX model with feedforward feature learning for multiclass categorization.

    Science.gov (United States)

    Li, Yinlin; Wu, Wei; Zhang, Bo; Li, Fengfu

    2015-01-01

    In recent years, the interdisciplinary research between neuroscience and computer vision has promoted the development in both fields. Many biologically inspired visual models are proposed, and among them, the Hierarchical Max-pooling model (HMAX) is a feedforward model mimicking the structures and functions of V1 to posterior inferotemporal (PIT) layer of the primate visual cortex, which could generate a series of position- and scale- invariant features. However, it could be improved with attention modulation and memory processing, which are two important properties of the primate visual cortex. Thus, in this paper, based on recent biological research on the primate visual cortex, we still mimic the first 100-150 ms of visual cognition to enhance the HMAX model, which mainly focuses on the unsupervised feedforward feature learning process. The main modifications are as follows: (1) To mimic the attention modulation mechanism of V1 layer, a bottom-up saliency map is computed in the S1 layer of the HMAX model, which can support the initial feature extraction for memory processing; (2) To mimic the learning, clustering and short-term memory to long-term memory conversion abilities of V2 and IT, an unsupervised iterative clustering method is used to learn clusters with multiscale middle level patches, which are taken as long-term memory; (3) Inspired by the multiple feature encoding mode of the primate visual cortex, information including color, orientation, and spatial position are encoded in different layers of the HMAX model progressively. By adding a softmax layer at the top of the model, multiclass categorization experiments can be conducted, and the results on Caltech101 show that the enhanced model with a smaller memory size exhibits higher accuracy than the original HMAX model, and could also achieve better accuracy than other unsupervised feature learning methods in multiclass categorization task.

  19. Unsupervised Language Acquisition

    Science.gov (United States)

    de Marcken, Carl

    1996-11-01

    This thesis presents a computational theory of unsupervised language acquisition, precisely defining procedures for learning language from ordinary spoken or written utterances, with no explicit help from a teacher. The theory is based heavily on concepts borrowed from machine learning and statistical estimation. In particular, learning takes place by fitting a stochastic, generative model of language to the evidence. Much of the thesis is devoted to explaining conditions that must hold for this general learning strategy to arrive at linguistically desirable grammars. The thesis introduces a variety of technical innovations, among them a common representation for evidence and grammars, and a learning strategy that separates the ``content'' of linguistic parameters from their representation. Algorithms based on it suffer from few of the search problems that have plagued other computational approaches to language acquisition. The theory has been tested on problems of learning vocabularies and grammars from unsegmented text and continuous speech, and mappings between sound and representations of meaning. It performs extremely well on various objective criteria, acquiring knowledge that causes it to assign almost exactly the same structure to utterances as humans do. This work has application to data compression, language modeling, speech recognition, machine translation, information retrieval, and other tasks that rely on either structural or stochastic descriptions of language.

  20. Investigation on IMCP based clustering in LTE-M communication for smart metering applications

    Directory of Open Access Journals (Sweden)

    Kartik Vishal Deshpande

    2017-06-01

    Full Text Available Machine to Machine (M2M is foreseen as an emerging technology for smart metering applications where devices communicate seamlessly for information transfer. The M2M communication makes use of long term evolution (LTE as its backbone network and it results in long-term evolution for machine type communication (LTE-M network. As huge number of M2M devices is to be handled by single eNB (evolved Node B, clustering is exploited for efficient processing of the network. This paper investigates the proposed Improved M2M Clustering Process (IMCP based clustering technique and it is compared with two well-known clustering algorithms, namely, Low Energy Adaptive Clustering Hierarchical (LEACH and Energy Aware Multihop Multipath Hierarchical (EAMMH techniques. Further, the IMCP algorithm is analyzed with two-tier and three-tier M2M systems for various mobility conditions. The proposed IMCP algorithm improves the last node death by 63.15% and 51.61% as compared to LEACH and EAMMH, respectively. Further, the average energy of each node in IMCP is increased by 89.85% and 81.15%, as compared to LEACH