Neural nets for job-shop scheduling, will they do the job?

NARCIS (Netherlands)

Rooda, J.E.; Willems, T.M.; Goodwin, G.C.; Evans, R.J.

1993-01-01

A neural net structure has been developed which is capable of solving deterministic jobshop scheduling problems, part of the large class of np-complete problems. The problem was translated in an integer linear-programming format which facilitated translation in an adequate neural net structure. Use

Bayesian Inference using Neural Net Likelihood Models for Protein Secondary Structure Prediction

Directory of Open Access Journals (Sweden)

Seong-Gon Kim

2011-06-01

Full Text Available Several techniques such as Neural Networks, Genetic Algorithms, Decision Trees and other statistical or heuristic methods have been used to approach the complex non-linear task of predicting Alpha-helicies, Beta-sheets and Turns of a proteins secondary structure in the past. This project introduces a new machine learning method by using an offline trained Multilayered Perceptrons (MLP as the likelihood models within a Bayesian Inference framework to predict secondary structures proteins. Varying window sizes are used to extract neighboring amino acid information and passed back and forth between the Neural Net models and the Bayesian Inference process until there is a convergence of the posterior secondary structure probability.

Deep learning with convolutional neural networks for EEG decoding and visualization.

Science.gov (United States)

Schirrmeister, Robin Tibor; Springenberg, Jost Tobias; Fiederer, Lukas Dominique Josef; Glasstetter, Martin; Eggensperger, Katharina; Tangermann, Michael; Hutter, Frank; Burgard, Wolfram; Ball, Tonio

2017-11-01

Deep learning with convolutional neural networks (deep ConvNets) has revolutionized computer vision through end-to-end learning, that is, learning from the raw data. There is increasing interest in using deep ConvNets for end-to-end EEG analysis, but a better understanding of how to design and train ConvNets for end-to-end EEG decoding and how to visualize the informative EEG features the ConvNets learn is still needed. Here, we studied deep ConvNets with a range of different architectures, designed for decoding imagined or executed tasks from raw EEG. Our results show that recent advances from the machine learning field, including batch normalization and exponential linear units, together with a cropped training strategy, boosted the deep ConvNets decoding performance, reaching at least as good performance as the widely used filter bank common spatial patterns (FBCSP) algorithm (mean decoding accuracies 82.1% FBCSP, 84.0% deep ConvNets). While FBCSP is designed to use spectral power modulations, the features used by ConvNets are not fixed a priori. Our novel methods for visualizing the learned features demonstrated that ConvNets indeed learned to use spectral power modulations in the alpha, beta, and high gamma frequencies, and proved useful for spatially mapping the learned features by revealing the topography of the causal contributions of features in different frequency bands to the decoding decision. Our study thus shows how to design and train ConvNets to decode task-related information from the raw EEG without handcrafted features and highlights the potential of deep ConvNets combined with advanced visualization techniques for EEG-based brain mapping. Hum Brain Mapp 38:5391-5420, 2017. © 2017 Wiley Periodicals, Inc. © 2017 The Authors Human Brain Mapping Published by Wiley Periodicals, Inc.

Deep learning with convolutional neural networks for EEG decoding and visualization

Science.gov (United States)

Springenberg, Jost Tobias; Fiederer, Lukas Dominique Josef; Glasstetter, Martin; Eggensperger, Katharina; Tangermann, Michael; Hutter, Frank; Burgard, Wolfram; Ball, Tonio

2017-01-01

Abstract Deep learning with convolutional neural networks (deep ConvNets) has revolutionized computer vision through end‐to‐end learning, that is, learning from the raw data. There is increasing interest in using deep ConvNets for end‐to‐end EEG analysis, but a better understanding of how to design and train ConvNets for end‐to‐end EEG decoding and how to visualize the informative EEG features the ConvNets learn is still needed. Here, we studied deep ConvNets with a range of different architectures, designed for decoding imagined or executed tasks from raw EEG. Our results show that recent advances from the machine learning field, including batch normalization and exponential linear units, together with a cropped training strategy, boosted the deep ConvNets decoding performance, reaching at least as good performance as the widely used filter bank common spatial patterns (FBCSP) algorithm (mean decoding accuracies 82.1% FBCSP, 84.0% deep ConvNets). While FBCSP is designed to use spectral power modulations, the features used by ConvNets are not fixed a priori. Our novel methods for visualizing the learned features demonstrated that ConvNets indeed learned to use spectral power modulations in the alpha, beta, and high gamma frequencies, and proved useful for spatially mapping the learned features by revealing the topography of the causal contributions of features in different frequency bands to the decoding decision. Our study thus shows how to design and train ConvNets to decode task‐related information from the raw EEG without handcrafted features and highlights the potential of deep ConvNets combined with advanced visualization techniques for EEG‐based brain mapping. Hum Brain Mapp 38:5391–5420, 2017. © 2017 Wiley Periodicals, Inc. PMID:28782865

Learning of N-layers neural network

Directory of Open Access Journals (Sweden)

Vladimír Konečný

2005-01-01

Full Text Available In the last decade we can observe increasing number of applications based on the Artificial Intelligence that are designed to solve problems from different areas of human activity. The reason why there is so much interest in these technologies is that the classical way of solutions does not exist or these technologies are not suitable because of their robustness. They are often used in applications like Business Intelligence that enable to obtain useful information for high-quality decision-making and to increase competitive advantage.One of the most widespread tools for the Artificial Intelligence are the artificial neural networks. Their high advantage is relative simplicity and the possibility of self-learning based on set of pattern situations.For the learning phase is the most commonly used algorithm back-propagation error (BPE. The base of BPE is the method minima of error function representing the sum of squared errors on outputs of neural net, for all patterns of the learning set. However, while performing BPE and in the first usage, we can find out that it is necessary to complete the handling of the learning factor by suitable method. The stability of the learning process and the rate of convergence depend on the selected method. In the article there are derived two functions: one function for the learning process management by the relative great error function value and the second function when the value of error function approximates to global minimum.The aim of the article is to introduce the BPE algorithm in compact matrix form for multilayer neural networks, the derivation of the learning factor handling method and the presentation of the results.

Adaptive Learning Rule for Hardware-based Deep Neural Networks Using Electronic Synapse Devices

OpenAIRE

Lim, Suhwan; Bae, Jong-Ho; Eum, Jai-Ho; Lee, Sungtae; Kim, Chul-Heung; Kwon, Dongseok; Park, Byung-Gook; Lee, Jong-Ho

2017-01-01

In this paper, we propose a learning rule based on a back-propagation (BP) algorithm that can be applied to a hardware-based deep neural network (HW-DNN) using electronic devices that exhibit discrete and limited conductance characteristics. This adaptive learning rule, which enables forward, backward propagation, as well as weight updates in hardware, is helpful during the implementation of power-efficient and high-speed deep neural networks. In simulations using a three-layer perceptron net...

Wet gas metering with the v-cone and neural nets

Energy Technology Data Exchange (ETDEWEB)

Toral, Haluk; Cai, Shiqian; Peters, Robert

2005-07-01

The paper presents analysis of extensive measurements taken at NEL, K-Lab and CEESI wet gas test loops. Differential and absolute pressure signals were sampled at high frequency across V-Cone meters. Turbulence characteristics of the flow captured in the sampled signals were characterized by pattern recognition techniques and related to the fractions and flow rates of individual phases. The sensitivity of over-reading to first and higher order features of the high frequency signals were investigated qualitatively. The sensitivities were quantified by means of the saliency test based on back propagating neural nets. A self contained wet gas meter based on neural net characterization of first and higher order features of the pressure, differential pressure and capacitance signals was proposed. Alternatively, a wet gas meter based on a neural net model of just pressure sensor inputs (based on currently available data) and liquid Froude number was shown to offer an accuracy of under 5% if the Froude number could be estimated with 25% accuracy. (author) (tk)

Neural net based determination of generator-shedding requirements in electric power systems

Energy Technology Data Exchange (ETDEWEB)

Djukanovic, M [Electrical Engineering Inst. ' Nikola Tesla' , Belgrade (Yugoslavia); Sobajic, D J; Pao, Y -H [Case Western Reserve Univ., Cleveland, OH (United States). Dept. of Electrical Engineering and Applied Physics Case Western Reserve Univ., Cleveland, OH (United States). Dept. of Computer Engineering and Science AI WARE Inc., Cleveland, OH (United States)

1992-09-01

This paper presents an application of artificial neural networks (ANN) in support of a decision-making process by power system operators directed towards the fast stabilisation of multi-machine systems. The proposed approach considers generator shedding as the most effective discrete supplementary control for improving the dynamic performance of faulted power systems and preventing instabilities. The sensitivity of the transient energy function (TEF) with respect to changes in the amount of dropped generation is used during the training phase of ANNs to assess the critical amount of generator shedding required to prevent the loss of synchronism. The learning capabilities of neural nets are used to establish complex mappings between fault information and the amount of generation to be shed, suggesting it as the control signal to the power system operator. (author)

Learning and Generalisation in Neural Networks with Local Preprocessing

OpenAIRE

Kutsia, Merab

2007-01-01

We study learning and generalisation ability of a specific two-layer feed-forward neural network and compare its properties to that of a simple perceptron. The input patterns are mapped nonlinearly onto a hidden layer, much larger than the input layer, and this mapping is either fixed or may result from an unsupervised learning process. Such preprocessing of initially uncorrelated random patterns results in the correlated patterns in the hidden layer. The hidden-to-output mapping of the net...

Construction of Neural Networks for Realization of Localized Deep Learning

Directory of Open Access Journals (Sweden)

Charles K. Chui

2018-05-01

Full Text Available The subject of deep learning has recently attracted users of machine learning from various disciplines, including: medical diagnosis and bioinformatics, financial market analysis and online advertisement, speech and handwriting recognition, computer vision and natural language processing, time series forecasting, and search engines. However, theoretical development of deep learning is still at its infancy. The objective of this paper is to introduce a deep neural network (also called deep-net approach to localized manifold learning, with each hidden layer endowed with a specific learning task. For the purpose of illustrations, we only focus on deep-nets with three hidden layers, with the first layer for dimensionality reduction, the second layer for bias reduction, and the third layer for variance reduction. A feedback component is also designed to deal with outliers. The main theoretical result in this paper is the order O(m-2s/(2s+d of approximation of the regression function with regularity s, in terms of the number m of sample points, where the (unknown manifold dimension d replaces the dimension D of the sampling (Euclidean space for shallow nets.

Artificial neural nets application in the cotton yarn industry

Directory of Open Access Journals (Sweden)

Gilberto Clóvis Antoneli

2016-06-01

Full Text Available The competitiveness in the yarn production sector has led companies to search for solutions to attain quality yarn at a low cost. Today, the difference between them, and thus the sector, is in the raw material, meaning processed cotton and its characteristics. There are many types of cotton with different characteristics due to its production region, harvest, storage and transportation. Yarn industries work with cotton mixtures, which makes it difficult to determine the quality of the yarn produced from the characteristics of the processed fibers. This study uses data from a conventional spinning, from a raw material made of 100% cotton, and presents a solution with artificial neural nets that determine the thread quality information, using the fibers’ characteristics values and settings of some process adjustments. In this solution a neural net of the type MultiLayer Perceptron with 11 entry neurons (8 characteristics of the fiber and 3 process adjustments, 7 output neurons (yarn quality and two types of training, Back propagation and Conjugate gradient descent. The selection and organization of the production data of the yarn industry of the cocamar® indústria de fios company are described, to apply the artificial neural nets developed. In the application of neural nets to determine yarn quality, one concludes that, although the ideal precision of absolute values is lacking, the presented solution represents an excellent tool to define yarn quality variations when modifying the raw material composition. The developed system enables a simulation to define the raw material percentage mixture to be processed in the plant using the information from the stocked cotton packs, thus obtaining a mixture that maintains the stability of the entire productive process.

Neural-net based real-time economic dispatch for thermal power plants

Energy Technology Data Exchange (ETDEWEB)

Djukanovic, M.; Milosevic, B. [Inst. Nikola Tesla, Belgrade (Yugoslavia). Dept. of Power Systems; Calovic, M. [Univ. of Belgrade (Yugoslavia). Dept. of Electrical Engineering; Sobajic, D.J. [Electric Power Research Inst., Palo Alto, CA (United States)

1996-12-01

This paper proposes the application of artificial neural networks to real-time optimal generation dispatch of thermal units. The approach can take into account the operational requirements and network losses. The proposed economic dispatch uses an artificial neural network (ANN) for generation of penalty factors, depending on the input generator powers and identified system load change. Then, a few additional iterations are performed within an iterative computation procedure for the solution of coordination equations, by using reference-bus penalty-factors derived from the Newton-Raphson load flow. A coordination technique for environmental and economic dispatch of pure thermal systems, based on the neural-net theory for simplified solution algorithms and improved man-machine interface is introduced. Numerical results on two test examples show that the proposed algorithm can efficiently and accurately develop optimal and feasible generator output trajectories, by applying neural-net forecasts of system load patterns.

Learning from neural control.

Science.gov (United States)

Wang, Cong; Hill, David J

2006-01-01

One of the amazing successes of biological systems is their ability to "learn by doing" and so adapt to their environment. In this paper, first, a deterministic learning mechanism is presented, by which an appropriately designed adaptive neural controller is capable of learning closed-loop system dynamics during tracking control to a periodic reference orbit. Among various neural network (NN) architectures, the localized radial basis function (RBF) network is employed. A property of persistence of excitation (PE) for RBF networks is established, and a partial PE condition of closed-loop signals, i.e., the PE condition of a regression subvector constructed out of the RBFs along a periodic state trajectory, is proven to be satisfied. Accurate NN approximation for closed-loop system dynamics is achieved in a local region along the periodic state trajectory, and a learning ability is implemented during a closed-loop feedback control process. Second, based on the deterministic learning mechanism, a neural learning control scheme is proposed which can effectively recall and reuse the learned knowledge to achieve closed-loop stability and improved control performance. The significance of this paper is that the presented deterministic learning mechanism and the neural learning control scheme provide elementary components toward the development of a biologically-plausible learning and control methodology. Simulation studies are included to demonstrate the effectiveness of the approach.

[A method of recognizing biology surface spectrum using cascade-connection artificial neural nets].

Science.gov (United States)

Shi, Wei-Jie; Yao, Yong; Zhang, Tie-Qiang; Meng, Xian-Jiang

2008-05-01

A method of recognizing the visible spectrum of micro-areas on the biological surface with cascade-connection artificial neural nets is presented in the present paper. The visible spectra of spots on apples' pericarp, ranging from 500 to 730 nm, were obtained with a fiber-probe spectrometer, and a new spectrum recognition system consisting of three-level cascade-connection neural nets was set up. The experiments show that the spectra of rotten, scar and bumped spot on an apple's pericarp can be recognized by the spectrum recognition system, and the recognition accuracy is higher than 85% even when noise level is 15%. The new recognition system overcomes the disadvantages of poor accuracy and poor anti-noise with the traditional system based on single cascade neural nets. Finally, a new method of expression of recognition results was proved. The method is based on the conception of degree of membership in fuzzing mathematics, and through it the recognition results can be expressed exactly and objectively.

A novel low-voltage low-power analogue VLSI implementation of neural networks with on-chip back-propagation learning

Science.gov (United States)

Carrasco, Manuel; Garde, Andres; Murillo, Pilar; Serrano, Luis

2005-06-01

In this paper a novel design and implementation of a VLSI Analogue Neural Net based on Multi-Layer Perceptron (MLP) with on-chip Back Propagation (BP) learning algorithm suitable for the resolution of classification problems is described. In order to implement a general and programmable analogue architecture, the design has been carried out in a hierarchical way. In this way the net has been divided in synapsis-blocks and neuron-blocks providing an easy method for the analysis. These blocks basically consist on simple cells, which are mainly, the activation functions (NAF), derivatives (DNAF), multipliers and weight update circuits. The analogue design is based on current-mode translinear techniques using MOS transistors working in the weak inversion region in order to reduce both the voltage supply and the power consumption. Moreover, with the purpose of minimizing the noise, offset and distortion of even order, the topologies are fully-differential and balanced. The circuit, named ANNE (Analogue Neural NEt), has been prototyped and characterized as a proof of concept on CMOS AMI-0.5A technology occupying a total area of 2.7mm2. The chip includes two versions of neural nets with on-chip BP learning algorithm, which are respectively a 2-1 and a 2-2-1 implementations. The proposed nets have been experimentally tested using supply voltages from 2.5V to 1.8V, which is suitable for single cell lithium-ion battery supply applications. Experimental results of both implementations included in ANNE exhibit a good performance on solving classification problems. These results have been compared with other proposed Analogue VLSI implementations of Neural Nets published in the literature demonstrating that our proposal is very efficient in terms of occupied area and power consumption.

Deep learning for steganalysis via convolutional neural networks

Science.gov (United States)

Qian, Yinlong; Dong, Jing; Wang, Wei; Tan, Tieniu

2015-03-01

Current work on steganalysis for digital images is focused on the construction of complex handcrafted features. This paper proposes a new paradigm for steganalysis to learn features automatically via deep learning models. We novelly propose a customized Convolutional Neural Network for steganalysis. The proposed model can capture the complex dependencies that are useful for steganalysis. Compared with existing schemes, this model can automatically learn feature representations with several convolutional layers. The feature extraction and classification steps are unified under a single architecture, which means the guidance of classification can be used during the feature extraction step. We demonstrate the effectiveness of the proposed model on three state-of-theart spatial domain steganographic algorithms - HUGO, WOW, and S-UNIWARD. Compared to the Spatial Rich Model (SRM), our model achieves comparable performance on BOSSbase and the realistic and large ImageNet database.

Modulated error diffusion CGHs for neural nets

Science.gov (United States)

Vermeulen, Pieter J. E.; Casasent, David P.

1990-05-01

New modulated error diffusion CGHs (computer generated holograms) for optical computing are considered. Specific attention is given to their use in optical matrix-vector, associative processor, neural net and optical interconnection architectures. We consider lensless CGH systems (many CGHs use an external Fourier transform (FT) lens), the Fresnel sampling requirements, the effects of finite CGH apertures (sample and hold inputs), dot size correction (for laser recorders), and new applications for this novel encoding method (that devotes attention to quantization noise effects).

White blood cells identification system based on convolutional deep neural learning networks.

Science.gov (United States)

Shahin, A I; Guo, Yanhui; Amin, K M; Sharawi, Amr A

2017-11-16

White blood cells (WBCs) differential counting yields valued information about human health and disease. The current developed automated cell morphology equipments perform differential count which is based on blood smear image analysis. Previous identification systems for WBCs consist of successive dependent stages; pre-processing, segmentation, feature extraction, feature selection, and classification. There is a real need to employ deep learning methodologies so that the performance of previous WBCs identification systems can be increased. Classifying small limited datasets through deep learning systems is a major challenge and should be investigated. In this paper, we propose a novel identification system for WBCs based on deep convolutional neural networks. Two methodologies based on transfer learning are followed: transfer learning based on deep activation features and fine-tuning of existed deep networks. Deep acrivation featues are extracted from several pre-trained networks and employed in a traditional identification system. Moreover, a novel end-to-end convolutional deep architecture called "WBCsNet" is proposed and built from scratch. Finally, a limited balanced WBCs dataset classification is performed through the WBCsNet as a pre-trained network. During our experiments, three different public WBCs datasets (2551 images) have been used which contain 5 healthy WBCs types. The overall system accuracy achieved by the proposed WBCsNet is (96.1%) which is more than different transfer learning approaches or even the previous traditional identification system. We also present features visualization for the WBCsNet activation which reflects higher response than the pre-trained activated one. a novel WBCs identification system based on deep learning theory is proposed and a high performance WBCsNet can be employed as a pre-trained network. Copyright © 2017. Published by Elsevier B.V.

Neural networks for aircraft control

Science.gov (United States)

Linse, Dennis

1990-01-01

Current research in Artificial Neural Networks indicates that networks offer some potential advantages in adaptation and fault tolerance. This research is directed at determining the possible applicability of neural networks to aircraft control. The first application will be to aircraft trim. Neural network node characteristics, network topology and operation, neural network learning and example histories using neighboring optimal control with a neural net are discussed.

Musical Audio Synthesis Using Autoencoding Neural Nets

OpenAIRE

Sarroff, Andy; Casey, Michael A.

2014-01-01

With an optimal network topology and tuning of hyperpa-\\ud rameters, artificial neural networks (ANNs) may be trained\\ud to learn a mapping from low level audio features to one\\ud or more higher-level representations. Such artificial neu-\\ud ral networks are commonly used in classification and re-\\ud gression settings to perform arbitrary tasks. In this work\\ud we suggest repurposing autoencoding neural networks as\\ud musical audio synthesizers. We offer an interactive musi-\\ud cal audio synt...

Deep Learning Neural Networks and Bayesian Neural Networks in Data Analysis

Directory of Open Access Journals (Sweden)

Chernoded Andrey

2017-01-01

Full Text Available Most of the modern analyses in high energy physics use signal-versus-background classification techniques of machine learning methods and neural networks in particular. Deep learning neural network is the most promising modern technique to separate signal and background and now days can be widely and successfully implemented as a part of physical analysis. In this article we compare Deep learning and Bayesian neural networks application as a classifiers in an instance of top quark analysis.

Vector control of wind turbine on the basis of the fuzzy selective neural net*

Science.gov (United States)

Engel, E. A.; Kovalev, I. V.; Engel, N. E.

2016-04-01

An article describes vector control of wind turbine based on fuzzy selective neural net. Based on the wind turbine system’s state, the fuzzy selective neural net tracks an maximum power point under random perturbations. Numerical simulations are accomplished to clarify the applicability and advantages of the proposed vector wind turbine’s control on the basis of the fuzzy selective neuronet. The simulation results show that the proposed intelligent control of wind turbine achieves real-time control speed and competitive performance, as compared to a classical control model with PID controllers based on traditional maximum torque control strategy.

Face recognition: Eigenface, elastic matching, and neural nets

International Nuclear Information System (INIS)

Zhang, J.; Lades, M.

1997-01-01

This paper is a comparative study of three recently proposed algorithms for face recognition: eigenface, autoassociation and classification neural nets, and elastic matching. After these algorithms were analyzed under a common statistical decision framework, they were evaluated experimentally on four individual data bases, each with a moderate subject size, and a combined data base with more than a hundred different subjects. Analysis and experimental results indicate that the eigenface algorithm, which is essentially a minimum distance classifier, works well when lighting variation is small. Its performance deteriorates significantly as lighting variation increases. The elastic matching algorithm, on the other hand, is insensitive to lighting, face position, and expression variations and therefore is more versatile. The performance of the autoassociation and classification nets is upper bounded by that of the eigenface but is more difficult to implement in practice

tf_unet: Generic convolutional neural network U-Net implementation in Tensorflow

Science.gov (United States)

Akeret, Joel; Chang, Chihway; Lucchi, Aurelien; Refregier, Alexandre

2016-11-01

tf_unet mitigates radio frequency interference (RFI) signals in radio data using a special type of Convolutional Neural Network, the U-Net, that enables the classification of clean signal and RFI signatures in 2D time-ordered data acquired from a radio telescope. The code is not tied to a specific segmentation and can be used, for example, to detect radio frequency interference (RFI) in radio astronomy or galaxies and stars in widefield imaging data. This U-Net implementation can outperform classical RFI mitigation algorithms.

Estimação do volume de árvores utilizando redes neurais artificiais Estimate of tree volume using artificial neural nets

Directory of Open Access Journals (Sweden)

Eric Bastos Gorgens

2009-12-01

Full Text Available Rede neural artificial consiste em um conjunto de unidades que contêm funções matemáticas, unidas por pesos. As redes são capazes de aprender, mediante modificação dos pesos sinápticos, e generalizar o aprendizado para outros arquivos desconhecidos. O projeto de redes neurais é composto por três etapas: pré-processamento, processamento e, por fim, pós-processamento dos dados. Um dos problemas clássicos que podem ser abordados por redes é a aproximação de funções. Nesse grupo, pode-se incluir a estimação do volume de árvores. Foram utilizados quatro arquiteturas diferentes, cinco pré-processamentos e duas funções de ativação. As redes que se apresentaram estatisticamente iguais aos dados observados também foram analisadas quanto ao resíduo e à distribuição dos volumes e comparadas com a estimação de volume pelo modelo de Schumacher e Hall. As redes neurais formadas por neurônios, cuja função de ativação era exponencial, apresentaram estimativas estatisticamente iguais aos dados observados. As redes treinadas com os dados normalizados pelo método da interpolação linear e equalizados tiveram melhor desempenho na estimação.The artificial neural network consists of a set of units containing mathematical functions connected by weights. Such nets are capable of learning by means of synaptic weight modification, generalizing learning for other unknown archives. The neural network project comprises three stages: pre-processing, processing and post-processing of data. One of the classical problems approached by networks is function approximation. Tree volume estimate can be included in this group. Four different architectures, five pre-processings and two activation functions were used. The nets which were statistically similar to the observed data were also analyzed in relation to residue and volume and compared to the volume estimate provided by the Schumacher and Hall equation. The neural nets formed by

Entropy Learning in Neural Network

Directory of Open Access Journals (Sweden)

Geok See Ng

2017-12-01

Full Text Available In this paper, entropy term is used in the learning phase of a neural network.  As learning progresses, more hidden nodes get into saturation.  The early creation of such hidden nodes may impair generalisation.  Hence entropy approach is proposed to dampen the early creation of such nodes.  The entropy learning also helps to increase the importance of relevant nodes while dampening the less important nodes.  At the end of learning, the less important nodes can then be eliminated to reduce the memory requirements of the neural network.

NiftyNet: a deep-learning platform for medical imaging.

Science.gov (United States)

Gibson, Eli; Li, Wenqi; Sudre, Carole; Fidon, Lucas; Shakir, Dzhoshkun I; Wang, Guotai; Eaton-Rosen, Zach; Gray, Robert; Doel, Tom; Hu, Yipeng; Whyntie, Tom; Nachev, Parashkev; Modat, Marc; Barratt, Dean C; Ourselin, Sébastien; Cardoso, M Jorge; Vercauteren, Tom

2018-05-01

Medical image analysis and computer-assisted intervention problems are increasingly being addressed with deep-learning-based solutions. Established deep-learning platforms are flexible but do not provide specific functionality for medical image analysis and adapting them for this domain of application requires substantial implementation effort. Consequently, there has been substantial duplication of effort and incompatible infrastructure developed across many research groups. This work presents the open-source NiftyNet platform for deep learning in medical imaging. The ambition of NiftyNet is to accelerate and simplify the development of these solutions, and to provide a common mechanism for disseminating research outputs for the community to use, adapt and build upon. The NiftyNet infrastructure provides a modular deep-learning pipeline for a range of medical imaging applications including segmentation, regression, image generation and representation learning applications. Components of the NiftyNet pipeline including data loading, data augmentation, network architectures, loss functions and evaluation metrics are tailored to, and take advantage of, the idiosyncracies of medical image analysis and computer-assisted intervention. NiftyNet is built on the TensorFlow framework and supports features such as TensorBoard visualization of 2D and 3D images and computational graphs by default. We present three illustrative medical image analysis applications built using NiftyNet infrastructure: (1) segmentation of multiple abdominal organs from computed tomography; (2) image regression to predict computed tomography attenuation maps from brain magnetic resonance images; and (3) generation of simulated ultrasound images for specified anatomical poses. The NiftyNet infrastructure enables researchers to rapidly develop and distribute deep learning solutions for segmentation, regression, image generation and representation learning applications, or extend the platform to new

Neural networks and statistical learning

CERN Document Server

Du, Ke-Lin

2014-01-01

Providing a broad but in-depth introduction to neural network and machine learning in a statistical framework, this book provides a single, comprehensive resource for study and further research. All the major popular neural network models and statistical learning approaches are covered with examples and exercises in every chapter to develop a practical working understanding of the content. Each of the twenty-five chapters includes state-of-the-art descriptions and important research results on the respective topics. The broad coverage includes the multilayer perceptron, the Hopfield network, associative memory models, clustering models and algorithms, the radial basis function network, recurrent neural networks, principal component analysis, nonnegative matrix factorization, independent component analysis, discriminant analysis, support vector machines, kernel methods, reinforcement learning, probabilistic and Bayesian networks, data fusion and ensemble learning, fuzzy sets and logic, neurofuzzy models, hardw...

Neural plasticity of development and learning.

Science.gov (United States)

Galván, Adriana

2010-06-01

Development and learning are powerful agents of change across the lifespan that induce robust structural and functional plasticity in neural systems. An unresolved question in developmental cognitive neuroscience is whether development and learning share the same neural mechanisms associated with experience-related neural plasticity. In this article, I outline the conceptual and practical challenges of this question, review insights gleaned from adult studies, and describe recent strides toward examining this topic across development using neuroimaging methods. I suggest that development and learning are not two completely separate constructs and instead, that they exist on a continuum. While progressive and regressive changes are central to both, the behavioral consequences associated with these changes are closely tied to the existing neural architecture of maturity of the system. Eventually, a deeper, more mechanistic understanding of neural plasticity will shed light on behavioral changes across development and, more broadly, about the underlying neural basis of cognition. (c) 2010 Wiley-Liss, Inc.

NIRFaceNet: A Convolutional Neural Network for Near-Infrared Face Identification

Directory of Open Access Journals (Sweden)

Min Peng

2016-10-01

Full Text Available Near-infrared (NIR face recognition has attracted increasing attention because of its advantage of illumination invariance. However, traditional face recognition methods based on NIR are designed for and tested in cooperative-user applications. In this paper, we present a convolutional neural network (CNN for NIR face recognition (specifically face identification in non-cooperative-user applications. The proposed NIRFaceNet is modified from GoogLeNet, but has a more compact structure designed specifically for the Chinese Academy of Sciences Institute of Automation (CASIA NIR database and can achieve higher identification rates with less training time and less processing time. The experimental results demonstrate that NIRFaceNet has an overall advantage compared to other methods in the NIR face recognition domain when image blur and noise are present. The performance suggests that the proposed NIRFaceNet method may be more suitable for non-cooperative-user applications.

Neural-net based coordinated stabilizing control for the exciter and governor loops of low head hydropower plants

Energy Technology Data Exchange (ETDEWEB)

Djukanovic, M.; Novicevic, M.; Dobrijevic, D.; Babic, B. [Electrical Engineering Inst. Nikola Tesla, Belgrade (Yugoslavia); Sobajic, D.J. [Electric Power Research Inst., Palo Alto, CA (United States); Pao, Y.H. [Case Western Reserve Univ., Cleveland, OH (United States)]|[AI WARE, Inc., Cleveland, OH (United States)

1995-12-01

This paper presents a design technique of a new adaptive optimal controller of the low head hydropower plant using artificial neural networks (ANN). The adaptive controller is to operate in real time to improve the generating unit transients through the exciter input, the guide vane position and the runner blade position. The new design procedure is based on self-organization and the predictive estimation capabilities of neural-nets implemented through the cluster-wise segmented associative memory scheme. The developed neural-net based controller (NNC) whose control signals are adjusted using the on-line measurements, can offer better damping effects for generator oscillations over a wide range of operating conditions than conventional controllers. Digital simulations of hydropower plant equipped with low head Kaplan turbine are performed and the comparisons of conventional excitation-governor control, state-space optimal control and neural-net based control are presented. Results obtained on the non-linear mathematical model demonstrate that the effects of the NNC closely agree with those obtained using the state-space multivariable discrete-time optimal controllers.

Supervised Learning with Complex-valued Neural Networks

CERN Document Server

Suresh, Sundaram; Savitha, Ramasamy

2013-01-01

Recent advancements in the field of telecommunications, medical imaging and signal processing deal with signals that are inherently time varying, nonlinear and complex-valued. The time varying, nonlinear characteristics of these signals can be effectively analyzed using artificial neural networks.  Furthermore, to efficiently preserve the physical characteristics of these complex-valued signals, it is important to develop complex-valued neural networks and derive their learning algorithms to represent these signals at every step of the learning process. This monograph comprises a collection of new supervised learning algorithms along with novel architectures for complex-valued neural networks. The concepts of meta-cognition equipped with a self-regulated learning have been known to be the best human learning strategy. In this monograph, the principles of meta-cognition have been introduced for complex-valued neural networks in both the batch and sequential learning modes. For applications where the computati...

MoleculeNet: a benchmark for molecular machine learning.

Science.gov (United States)

Wu, Zhenqin; Ramsundar, Bharath; Feinberg, Evan N; Gomes, Joseph; Geniesse, Caleb; Pappu, Aneesh S; Leswing, Karl; Pande, Vijay

2018-01-14

Molecular machine learning has been maturing rapidly over the last few years. Improved methods and the presence of larger datasets have enabled machine learning algorithms to make increasingly accurate predictions about molecular properties. However, algorithmic progress has been limited due to the lack of a standard benchmark to compare the efficacy of proposed methods; most new algorithms are benchmarked on different datasets making it challenging to gauge the quality of proposed methods. This work introduces MoleculeNet, a large scale benchmark for molecular machine learning. MoleculeNet curates multiple public datasets, establishes metrics for evaluation, and offers high quality open-source implementations of multiple previously proposed molecular featurization and learning algorithms (released as part of the DeepChem open source library). MoleculeNet benchmarks demonstrate that learnable representations are powerful tools for molecular machine learning and broadly offer the best performance. However, this result comes with caveats. Learnable representations still struggle to deal with complex tasks under data scarcity and highly imbalanced classification. For quantum mechanical and biophysical datasets, the use of physics-aware featurizations can be more important than choice of particular learning algorithm.

NeMO-Net & Fluid Lensing: The Neural Multi-Modal Observation & Training Network for Global Coral Reef Assessment Using Fluid Lensing Augmentation of NASA EOS Data

Science.gov (United States)

Chirayath, Ved

2018-01-01

We present preliminary results from NASA NeMO-Net, the first neural multi-modal observation and training network for global coral reef assessment. NeMO-Net is an open-source deep convolutional neural network (CNN) and interactive active learning training software in development which will assess the present and past dynamics of coral reef ecosystems. NeMO-Net exploits active learning and data fusion of mm-scale remotely sensed 3D images of coral reefs captured using fluid lensing with the NASA FluidCam instrument, presently the highest-resolution remote sensing benthic imaging technology capable of removing ocean wave distortion, as well as hyperspectral airborne remote sensing data from the ongoing NASA CORAL mission and lower-resolution satellite data to determine coral reef ecosystem makeup globally at unprecedented spatial and temporal scales. Aquatic ecosystems, particularly coral reefs, remain quantitatively misrepresented by low-resolution remote sensing as a result of refractive distortion from ocean waves, optical attenuation, and remoteness. Machine learning classification of coral reefs using FluidCam mm-scale 3D data show that present satellite and airborne remote sensing techniques poorly characterize coral reef percent living cover, morphology type, and species breakdown at the mm, cm, and meter scales. Indeed, current global assessments of coral reef cover and morphology classification based on km-scale satellite data alone can suffer from segmentation errors greater than 40%, capable of change detection only on yearly temporal scales and decameter spatial scales, significantly hindering our understanding of patterns and processes in marine biodiversity at a time when these ecosystems are experiencing unprecedented anthropogenic pressures, ocean acidification, and sea surface temperature rise. NeMO-Net leverages our augmented machine learning algorithm that demonstrates data fusion of regional FluidCam (mm, cm-scale) airborne remote sensing with

Neural net prediction of tokamak plasma disruptions

International Nuclear Information System (INIS)

Hernandez, J.V.; Lin, Z.; Horton, W.; McCool, S.C.

1994-10-01

The computation based on neural net algorithms in predicting minor and major disruptions in TEXT tokamak discharges has been performed. Future values of the fluctuating magnetic signal are predicted based on L past values of the magnetic fluctuation signal, measured by a single Mirnov coil. The time step used (= 0.04ms) corresponds to the experimental data sampling rate. Two kinds of approaches are adopted for the task, the contiguous future prediction and the multi-timescale prediction. Results are shown for comparison. Both networks are trained through the back-propagation algorithm with inertial terms. The degree of this success indicates that the magnetic fluctuations associated with tokamak disruptions may be characterized by a relatively low-dimensional dynamical system

Generalized Net Model of the Cognitive and Neural Algorithm for Adaptive Resonance Theory 1

Directory of Open Access Journals (Sweden)

Todor Petkov

2013-12-01

Full Text Available The artificial neural networks are inspired by biological properties of human and animal brains. One of the neural networks type is called ART [4]. The abbreviation of ART stands for Adaptive Resonance Theory that has been invented by Stephen Grossberg in 1976 [5]. ART represents a family of Neural Networks. It is a cognitive and neural theory that describes how the brain autonomously learns to categorize, recognize and predict objects and events in the changing world. In this paper we introduce a GN model that represent ART1 Neural Network learning algorithm [1]. The purpose of this model is to explain when the input vector will be clustered or rejected among all nodes by the network. It can also be used for explanation and optimization of ART1 learning algorithm.

Global reinforcement training of CrossNets

Science.gov (United States)

Ma, Xiaolong

2007-10-01

Hybrid "CMOL" integrated circuits, incorporating advanced CMOS devices for neural cell bodies, nanowires as axons and dendrites, and latching switches as synapses, may be used for the hardware implementation of extremely dense (107 cells and 1012 synapses per cm2) neuromorphic networks, operating up to 10 6 times faster than their biological prototypes. We are exploring several "Cross- Net" architectures that accommodate the limitations imposed by CMOL hardware and should allow effective training of the networks without a direct external access to individual synapses. Our studies have show that CrossNets based on simple (two-terminal) crosspoint devices can work well in at least two modes: as Hop-field networks for associative memory and multilayer perceptrons for classification tasks. For more intelligent tasks (such as robot motion control or complex games), which do not have "examples" for supervised learning, more advanced training methods such as the global reinforcement learning are necessary. For application of global reinforcement training algorithms to CrossNets, we have extended Williams's REINFORCE learning principle to a more general framework and derived several learning rules that are more suitable for CrossNet hardware implementation. The results of numerical experiments have shown that these new learning rules can work well for both classification tasks and reinforcement tasks such as the cartpole balancing control problem. Some limitations imposed by the CMOL hardware need to be carefully addressed for the the successful application of in situ reinforcement training to CrossNets.

Transfer Learning with Convolutional Neural Networks for Classification of Abdominal Ultrasound Images.

Science.gov (United States)

Cheng, Phillip M; Malhi, Harshawn S

2017-04-01

The purpose of this study is to evaluate transfer learning with deep convolutional neural networks for the classification of abdominal ultrasound images. Grayscale images from 185 consecutive clinical abdominal ultrasound studies were categorized into 11 categories based on the text annotation specified by the technologist for the image. Cropped images were rescaled to 256 × 256 resolution and randomized, with 4094 images from 136 studies constituting the training set, and 1423 images from 49 studies constituting the test set. The fully connected layers of two convolutional neural networks based on CaffeNet and VGGNet, previously trained on the 2012 Large Scale Visual Recognition Challenge data set, were retrained on the training set. Weights in the convolutional layers of each network were frozen to serve as fixed feature extractors. Accuracy on the test set was evaluated for each network. A radiologist experienced in abdominal ultrasound also independently classified the images in the test set into the same 11 categories. The CaffeNet network classified 77.3% of the test set images accurately (1100/1423 images), with a top-2 accuracy of 90.4% (1287/1423 images). The larger VGGNet network classified 77.9% of the test set accurately (1109/1423 images), with a top-2 accuracy of VGGNet was 89.7% (1276/1423 images). The radiologist classified 71.7% of the test set images correctly (1020/1423 images). The differences in classification accuracies between both neural networks and the radiologist were statistically significant (p convolutional neural networks may be used to construct effective classifiers for abdominal ultrasound images.

Neural network error correction for solving coupled ordinary differential equations

Science.gov (United States)

Shelton, R. O.; Darsey, J. A.; Sumpter, B. G.; Noid, D. W.

1992-01-01

A neural network is presented to learn errors generated by a numerical algorithm for solving coupled nonlinear differential equations. The method is based on using a neural network to correctly learn the error generated by, for example, Runge-Kutta on a model molecular dynamics (MD) problem. The neural network programs used in this study were developed by NASA. Comparisons are made for training the neural network using backpropagation and a new method which was found to converge with fewer iterations. The neural net programs, the MD model and the calculations are discussed.

Unfolding code for neutron spectrometry based on neural nets technology

International Nuclear Information System (INIS)

Ortiz R, J. M.; Vega C, H. R.

2012-10-01

The most delicate part of neutron spectrometry, is the unfolding process. The derivation of the spectral information is not simple because the unknown is not given directly as a result of the measurements. The drawbacks associated with traditional unfolding procedures have motivated the need of complementary approaches. Novel methods based on Artificial Neural Networks have been widely investigated. In this work, a neutron spectrum unfolding code based on neural nets technology is presented. This unfolding code called Neutron Spectrometry and Dosimetry by means of Artificial Neural Networks was designed in a graphical interface under LabVIEW programming environment. The core of the code is an embedded neural network architecture, previously optimized by the R obust Design of Artificial Neural Networks Methodology . The main features of the code are: is easy to use, friendly and intuitive to the user. This code was designed for a Bonner Sphere System based on a 6 Lil(Eu) neutron detector and a response matrix expressed in 60 energy bins taken from an International Atomic Energy Agency compilation. The main feature of the code is that as entrance data, only seven rate counts measurement with a Bonner spheres spectrometer are required for simultaneously unfold the 60 energy bins of the neutron spectrum and to calculate 15 dosimetric quantities, for radiation protection porpoises. This code generates a full report in html format with all relevant information. (Author)

Neural net generated seismic facies map and attribute facies map

International Nuclear Information System (INIS)

Addy, S.K.; Neri, P.

1998-01-01

The usefulness of 'seismic facies maps' in the analysis of an Upper Wilcox channel system in a 3-D survey shot by CGG in 1995 in Lavaca county in south Texas was discussed. A neural net-generated seismic facies map is a quick hydrocarbon exploration tool that can be applied regionally as well as on a prospect scale. The new technology is used to classify a constant interval parallel to a horizon in a 3-D seismic volume based on the shape of the wiggle traces using a neural network technology. The tool makes it possible to interpret sedimentary features of a petroleum deposit. The same technology can be used in regional mapping by making 'attribute facies maps' in which various forms of amplitude attributes, phase attributes or frequency attributes can be used

Assessment of the expected construction company’s net profit using neural network and multiple regression models

Directory of Open Access Journals (Sweden)

H.H. Mohamad

2013-09-01

This research aims to develop a mathematical model for assessing the expected net profit of any construction company. To achieve the research objective, four steps were performed. First, the main factors affecting firms’ net profit were identified. Second, pertinent data regarding the net profit factors were collected. Third, two different net profit models were developed using the Multiple Regression (MR and the Neural Network (NN techniques. The validity of the proposed models was also investigated. Finally, the results of both MR and NN models were compared to investigate the predictive capabilities of the two models.

NeMO-Net - The Neural Multi-Modal Observation & Training Network for Global Coral Reef Assessment

Science.gov (United States)

Li, A. S. X.; Chirayath, V.; Segal-Rosenhaimer, M.; Das, K.

2017-12-01

In the past decade, coral reefs worldwide have experienced unprecedented stresses due to climate change, ocean acidification, and anthropomorphic pressures, instigating massive bleaching and die-off of these fragile and diverse ecosystems. Furthermore, remote sensing of these shallow marine habitats is hindered by ocean wave distortion, refraction and optical attenuation, leading invariably to data products that are often of low resolution and signal-to-noise (SNR) ratio. However, recent advances in UAV and Fluid Lensing technology have allowed us to capture multispectral 3D imagery of these systems at sub-cm scales from above the water surface, giving us an unprecedented view of their growth and decay. Exploiting the fine-scaled features of these datasets, machine learning methods such as MAP, PCA, and SVM can not only accurately classify the living cover and morphology of these reef systems (below 8% error), but are also able to map the spectral space between airborne and satellite imagery, augmenting and improving the classification accuracy of previously low-resolution datasets.We are currently implementing NeMO-Net, the first open-source deep convolutional neural network (CNN) and interactive active learning and training software to accurately assess the present and past dynamics of coral reef ecosystems through determination of percent living cover and morphology. NeMO-Net will be built upon the QGIS platform to ingest UAV, airborne and satellite datasets from various sources and sensor capabilities, and through data-fusion determine the coral reef ecosystem makeup globally at unprecedented spatial and temporal scales. To achieve this, we will exploit virtual data augmentation, the use of semi-supervised learning, and active learning through a tablet platform allowing for users to manually train uncertain or difficult to classify datasets. The project will make use of Python's extensive libraries for machine learning, as well as extending integration to GPU

Exemplar-based optical neural net classifier for color pattern recognition

Science.gov (United States)

Yu, Francis T. S.; Uang, Chii-Maw; Yang, Xiangyang

1992-10-01

We present a color exemplar-based neural network that can be used as an optimum image classifier or an associative memory. Color decomposition and composition technique is used for constructing the polychromatic interconnection weight matrix (IWM). The Hamming net algorithm is modified to relax the dynamic range requirement of the spatial light modulator and to reduce the number of iteration cycles in the winner-take-all layer. Computer simulation results demonstrated the feasibility of this approach

SupportNet: a novel incremental learning framework through deep learning and support data

KAUST Repository

Li, Yu; Li, Zhongxiao; Ding, Lizhong; Hu, Yuhui; Chen, Wei; Gao, Xin

2018-01-01

Motivation: In most biological data sets, the amount of data is regularly growing and the number of classes is continuously increasing. To deal with the new data from the new classes, one approach is to train a classification model, e.g., a deep learning model, from scratch based on both old and new data. This approach is highly computationally costly and the extracted features are likely very different from the ones extracted by the model trained on the old data alone, which leads to poor model robustness. Another approach is to fine tune the trained model from the old data on the new data. However, this approach often does not have the ability to learn new knowledge without forgetting the previously learned knowledge, which is known as the catastrophic forgetting problem. To our knowledge, this problem has not been studied in the field of bioinformatics despite its existence in many bioinformatic problems. Results: Here we propose a novel method, SupportNet, to solve the catastrophic forgetting problem efficiently and effectively. SupportNet combines the strength of deep learning and support vector machine (SVM), where SVM is used to identify the support data from the old data, which are fed to the deep learning model together with the new data for further training so that the model can review the essential information of the old data when learning the new information. Two powerful consolidation regularizers are applied to ensure the robustness of the learned model. Comprehensive experiments on various tasks, including enzyme function prediction, subcellular structure classification and breast tumor classification, show that SupportNet drastically outperforms the state-of-the-art incremental learning methods and reaches similar performance as the deep learning model trained from scratch on both old and new data. Availability: Our program is accessible at: \\url{https://github.com/lykaust15/SupportNet}.

SupportNet: a novel incremental learning framework through deep learning and support data

KAUST Repository

Li, Yu

2018-05-08

Motivation: In most biological data sets, the amount of data is regularly growing and the number of classes is continuously increasing. To deal with the new data from the new classes, one approach is to train a classification model, e.g., a deep learning model, from scratch based on both old and new data. This approach is highly computationally costly and the extracted features are likely very different from the ones extracted by the model trained on the old data alone, which leads to poor model robustness. Another approach is to fine tune the trained model from the old data on the new data. However, this approach often does not have the ability to learn new knowledge without forgetting the previously learned knowledge, which is known as the catastrophic forgetting problem. To our knowledge, this problem has not been studied in the field of bioinformatics despite its existence in many bioinformatic problems. Results: Here we propose a novel method, SupportNet, to solve the catastrophic forgetting problem efficiently and effectively. SupportNet combines the strength of deep learning and support vector machine (SVM), where SVM is used to identify the support data from the old data, which are fed to the deep learning model together with the new data for further training so that the model can review the essential information of the old data when learning the new information. Two powerful consolidation regularizers are applied to ensure the robustness of the learned model. Comprehensive experiments on various tasks, including enzyme function prediction, subcellular structure classification and breast tumor classification, show that SupportNet drastically outperforms the state-of-the-art incremental learning methods and reaches similar performance as the deep learning model trained from scratch on both old and new data. Availability: Our program is accessible at: \\\\url{https://github.com/lykaust15/SupportNet}.

Learning in Artificial Neural Systems

Science.gov (United States)

Matheus, Christopher J.; Hohensee, William E.

1987-01-01

This paper presents an overview and analysis of learning in Artificial Neural Systems (ANS's). It begins with a general introduction to neural networks and connectionist approaches to information processing. The basis for learning in ANS's is then described, and compared with classical Machine learning. While similar in some ways, ANS learning deviates from tradition in its dependence on the modification of individual weights to bring about changes in a knowledge representation distributed across connections in a network. This unique form of learning is analyzed from two aspects: the selection of an appropriate network architecture for representing the problem, and the choice of a suitable learning rule capable of reproducing the desired function within the given network. The various network architectures are classified, and then identified with explicit restrictions on the types of functions they are capable of representing. The learning rules, i.e., algorithms that specify how the network weights are modified, are similarly taxonomized, and where possible, the limitations inherent to specific classes of rules are outlined.

Shaping the learning curve: epigenetic dynamics in neural plasticity

Directory of Open Access Journals (Sweden)

Zohar Ziv Bronfman

2014-07-01

Full Text Available A key characteristic of learning and neural plasticity is state-dependent acquisition dynamics reflected by the non-linear learning curve that links increase in learning with practice. Here we propose that the manner by which epigenetic states of individual cells change during learning contributes to the shape of the neural and behavioral learning curve. We base our suggestion on recent studies showing that epigenetic mechanisms such as DNA methylation, histone acetylation and RNA-mediated gene regulation are intimately involved in the establishment and maintenance of long-term neural plasticity, reflecting specific learning-histories and influencing future learning. Our model, which is the first to suggest a dynamic molecular account of the shape of the learning curve, leads to several testable predictions regarding the link between epigenetic dynamics at the promoter, gene-network and neural-network levels. This perspective opens up new avenues for therapeutic interventions in neurological pathologies.

Real-time applications of neural nets

International Nuclear Information System (INIS)

Spencer, J.E.

1989-05-01

Producing, accelerating and colliding very high power, low emittance beams for long periods is a formidable problem in real-time control. As energy has grown exponentially in time so has the complexity of the machines and their control systems. Similar growth rates have occurred in many areas, e.g., improved integrated circuits have been paid for with comparable increases in complexity. However, in this case, reliability, capability and cost have improved due to reduced size, high production and increased integration which allow various kinds of feedback. In contrast, most large complex systems (LCS) are perceived to lack such possibilities because only one copy is made. Neural nets, as a metaphor for LCS, suggest ways to circumvent such limitations. It is argued that they are logically equivalent to multi-loop feedback/forward control of faulty systems. While complimentary to AI, they mesh nicely with characteristics desired for real-time systems. Such issues are considered, examples given and possibilities discussed. 21 refs., 6 figs

Real-time applications of neural nets

International Nuclear Information System (INIS)

Spencer, J.E.

1989-01-01

Producing, accelerating and colliding very high power, low emittance beams for long periods is a formidable problem in real-time control. As energy has grown exponentially in time so has the complexity of the machines and their control systems. Similar growth rates have occurred in many areas e.g. improved integrated circuits have been paid for with comparable increases in complexity. However, in this case, reliability, capability and cost have improved due to reduced size, high production and increased integration which allow various kinds of feedback. In contrast, most large complex systems (LCS) are perceived to lack such possibilities because only one copy is made. Neural nets, as a metaphor for LCS, suggest ways to circumvent such limitations. It is argued that they are logically equivalent to multi-loop feedback/forward control of faulty systems. While complimentary to AI, they mesh nicely with characteristics desired for real-time systems. In this paper, such issues are considered, examples given and possibilities discussed

Real-time applications of neural nets

Energy Technology Data Exchange (ETDEWEB)

Spencer, J.E.

1989-05-01

Producing, accelerating and colliding very high power, low emittance beams for long periods is a formidable problem in real-time control. As energy has grown exponentially in time so has the complexity of the machines and their control systems. Similar growth rates have occurred in many areas, e.g., improved integrated circuits have been paid for with comparable increases in complexity. However, in this case, reliability, capability and cost have improved due to reduced size, high production and increased integration which allow various kinds of feedback. In contrast, most large complex systems (LCS) are perceived to lack such possibilities because only one copy is made. Neural nets, as a metaphor for LCS, suggest ways to circumvent such limitations. It is argued that they are logically equivalent to multi-loop feedback/forward control of faulty systems. While complimentary to AI, they mesh nicely with characteristics desired for real-time systems. Such issues are considered, examples given and possibilities discussed. 21 refs., 6 figs.

Unfolding code for neutron spectrometry based on neural nets technology

Energy Technology Data Exchange (ETDEWEB)

Ortiz R, J. M.; Vega C, H. R., E-mail: morvymm@yahoo.com.mx [Universidad Autonoma de Zacatecas, Unidad Academica de Ingenieria Electrica, Apdo. Postal 336, 98000 Zacatecas (Mexico)

2012-10-15

The most delicate part of neutron spectrometry, is the unfolding process. The derivation of the spectral information is not simple because the unknown is not given directly as a result of the measurements. The drawbacks associated with traditional unfolding procedures have motivated the need of complementary approaches. Novel methods based on Artificial Neural Networks have been widely investigated. In this work, a neutron spectrum unfolding code based on neural nets technology is presented. This unfolding code called Neutron Spectrometry and Dosimetry by means of Artificial Neural Networks was designed in a graphical interface under LabVIEW programming environment. The core of the code is an embedded neural network architecture, previously optimized by the {sup R}obust Design of Artificial Neural Networks Methodology{sup .} The main features of the code are: is easy to use, friendly and intuitive to the user. This code was designed for a Bonner Sphere System based on a {sup 6}Lil(Eu) neutron detector and a response matrix expressed in 60 energy bins taken from an International Atomic Energy Agency compilation. The main feature of the code is that as entrance data, only seven rate counts measurement with a Bonner spheres spectrometer are required for simultaneously unfold the 60 energy bins of the neutron spectrum and to calculate 15 dosimetric quantities, for radiation protection porpoises. This code generates a full report in html format with all relevant information. (Author)

Bootstrapped neural nets versus regression kriging in the digital mapping of pedological attributes: the automatic and time-consuming perspectives

Science.gov (United States)

Langella, Giuliano; Basile, Angelo; Bonfante, Antonello; Manna, Piero; Terribile, Fabio

2013-04-01

Digital soil mapping procedures are widespread used to build two-dimensional continuous maps about several pedological attributes. Our work addressed a regression kriging (RK) technique and a bootstrapped artificial neural network approach in order to evaluate and compare (i) the accuracy of prediction, (ii) the susceptibility of being included in automatic engines (e.g. to constitute web processing services), and (iii) the time cost needed for calibrating models and for making predictions. Regression kriging is maybe the most widely used geostatistical technique in the digital soil mapping literature. Here we tried to apply the EBLUP regression kriging as it is deemed to be the most statistically sound RK flavor by pedometricians. An unusual multi-parametric and nonlinear machine learning approach was accomplished, called BAGAP (Bootstrap aggregating Artificial neural networks with Genetic Algorithms and Principal component regression). BAGAP combines a selected set of weighted neural nets having specified characteristics to yield an ensemble response. The purpose of applying these two particular models is to ascertain whether and how much a more cumbersome machine learning method could be much promising in making more accurate/precise predictions. Being aware of the difficulty to handle objects based on EBLUP-RK as well as BAGAP when they are embedded in environmental applications, we explore the susceptibility of them in being wrapped within Web Processing Services. Two further kinds of aspects are faced for an exhaustive evaluation and comparison: automaticity and time of calculation with/without high performance computing leverage.

A theory of causal learning in children: causal maps and Bayes nets.

Science.gov (United States)

Gopnik, Alison; Glymour, Clark; Sobel, David M; Schulz, Laura E; Kushnir, Tamar; Danks, David

2004-01-01

The authors outline a cognitive and computational account of causal learning in children. They propose that children use specialized cognitive systems that allow them to recover an accurate "causal map" of the world: an abstract, coherent, learned representation of the causal relations among events. This kind of knowledge can be perspicuously understood in terms of the formalism of directed graphical causal models, or Bayes nets. Children's causal learning and inference may involve computations similar to those for learning causal Bayes nets and for predicting with them. Experimental results suggest that 2- to 4-year-old children construct new causal maps and that their learning is consistent with the Bayes net formalism.

SchNet - A deep learning architecture for molecules and materials

Science.gov (United States)

Schütt, K. T.; Sauceda, H. E.; Kindermans, P.-J.; Tkatchenko, A.; Müller, K.-R.

2018-06-01

Deep learning has led to a paradigm shift in artificial intelligence, including web, text, and image search, speech recognition, as well as bioinformatics, with growing impact in chemical physics. Machine learning, in general, and deep learning, in particular, are ideally suitable for representing quantum-mechanical interactions, enabling us to model nonlinear potential-energy surfaces or enhancing the exploration of chemical compound space. Here we present the deep learning architecture SchNet that is specifically designed to model atomistic systems by making use of continuous-filter convolutional layers. We demonstrate the capabilities of SchNet by accurately predicting a range of properties across chemical space for molecules and materials, where our model learns chemically plausible embeddings of atom types across the periodic table. Finally, we employ SchNet to predict potential-energy surfaces and energy-conserving force fields for molecular dynamics simulations of small molecules and perform an exemplary study on the quantum-mechanical properties of C20-fullerene that would have been infeasible with regular ab initio molecular dynamics.

Pattern recognition neural-net by spatial mapping of biology visual field

Science.gov (United States)

Lin, Xin; Mori, Masahiko

2000-05-01

The method of spatial mapping in biology vision field is applied to artificial neural networks for pattern recognition. By the coordinate transform that is called the complex-logarithm mapping and Fourier transform, the input images are transformed into scale- rotation- and shift- invariant patterns, and then fed into a multilayer neural network for learning and recognition. The results of computer simulation and an optical experimental system are described.

EnzyNet: enzyme classification using 3D convolutional neural networks on spatial representation.

Science.gov (United States)

Amidi, Afshine; Amidi, Shervine; Vlachakis, Dimitrios; Megalooikonomou, Vasileios; Paragios, Nikos; Zacharaki, Evangelia I

2018-01-01

During the past decade, with the significant progress of computational power as well as ever-rising data availability, deep learning techniques became increasingly popular due to their excellent performance on computer vision problems. The size of the Protein Data Bank (PDB) has increased more than 15-fold since 1999, which enabled the expansion of models that aim at predicting enzymatic function via their amino acid composition. Amino acid sequence, however, is less conserved in nature than protein structure and therefore considered a less reliable predictor of protein function. This paper presents EnzyNet, a novel 3D convolutional neural networks classifier that predicts the Enzyme Commission number of enzymes based only on their voxel-based spatial structure. The spatial distribution of biochemical properties was also examined as complementary information. The two-layer architecture was investigated on a large dataset of 63,558 enzymes from the PDB and achieved an accuracy of 78.4% by exploiting only the binary representation of the protein shape. Code and datasets are available at https://github.com/shervinea/enzynet.

StarNet: An application of deep learning in the analysis of stellar spectra

Science.gov (United States)

Kielty, Collin; Bialek, Spencer; Fabbro, Sebastien; Venn, Kim; O'Briain, Teaghan; Jahandar, Farbod; Monty, Stephanie

2018-06-01

In an era when spectroscopic surveys are capable of collecting spectra for hundreds of thousands of stars, fast and efficient analysis methods are required to maximize scientific impact. These surveys provide a homogeneous database of stellar spectra that are ideal for machine learning applications. In this poster, we present StarNet: a convolutional neural network model applied to the analysis of both SDSS-III APOGEE DR13 and synthetic stellar spectra. When trained on synthetic spectra alone, the calculated stellar parameters (temperature, surface gravity, and metallicity) are of excellent precision and accuracy for both APOGEE data and synthetic data, over a wide range of signal-to-noise ratios. While StarNet was developed using the APOGEE observed spectra and corresponding ASSeT synthetic grid, we suggest that this technique is applicable to other spectral resolutions, spectral surveys, and wavelength regimes. As a demonstration of this, we present a StarNet model trained on lower resolution, R=6000, IR synthetic spectra, describing the spectra delivered by Gemini/NIFS and the forthcoming Gemini/GIRMOS instrument (PI Sivanandam, UToronto). Preliminary results suggest that the stellar parameters determined from this low resolution StarNet model are comparable in precision to the high-resolution APOGEE results. The success of StarNet at lower resolution can be attributed to (1) a large training set of synthetic spectra (N ~200,000) with a priori stellar labels, and (2) the use of the entire spectrum in the solution rather than a few weighted windows, which are common methods in other spectral analysis tools (e.g. FERRE or The Cannon). Remaining challenges in our StarNet applications include rectification, continuum normalization, and wavelength coverage. Solutions to these problems could be used to guide decisions made in the development of future spectrographs, spectroscopic surveys, and data reduction pipelines, such as for the future MSE.

ClimateNet: A Machine Learning dataset for Climate Science Research

Science.gov (United States)

Prabhat, M.; Biard, J.; Ganguly, S.; Ames, S.; Kashinath, K.; Kim, S. K.; Kahou, S.; Maharaj, T.; Beckham, C.; O'Brien, T. A.; Wehner, M. F.; Williams, D. N.; Kunkel, K.; Collins, W. D.

2017-12-01

Deep Learning techniques have revolutionized commercial applications in Computer vision, speech recognition and control systems. The key for all of these developments was the creation of a curated, labeled dataset ImageNet, for enabling multiple research groups around the world to develop methods, benchmark performance and compete with each other. The success of Deep Learning can be largely attributed to the broad availability of this dataset. Our empirical investigations have revealed that Deep Learning is similarly poised to benefit the task of pattern detection in climate science. Unfortunately, labeled datasets, a key pre-requisite for training, are hard to find. Individual research groups are typically interested in specialized weather patterns, making it hard to unify, and share datasets across groups and institutions. In this work, we are proposing ClimateNet: a labeled dataset that provides labeled instances of extreme weather patterns, as well as associated raw fields in model and observational output. We develop a schema in NetCDF to enumerate weather pattern classes/types, store bounding boxes, and pixel-masks. We are also working on a TensorFlow implementation to natively import such NetCDF datasets, and are providing a reference convolutional architecture for binary classification tasks. Our hope is that researchers in Climate Science, as well as ML/DL, will be able to use (and extend) ClimateNet to make rapid progress in the application of Deep Learning for Climate Science research.

QML-AiNet: An immune network approach to learning qualitative differential equation models.

Science.gov (United States)

Pang, Wei; Coghill, George M

2015-02-01

In this paper, we explore the application of Opt-AiNet, an immune network approach for search and optimisation problems, to learning qualitative models in the form of qualitative differential equations. The Opt-AiNet algorithm is adapted to qualitative model learning problems, resulting in the proposed system QML-AiNet. The potential of QML-AiNet to address the scalability and multimodal search space issues of qualitative model learning has been investigated. More importantly, to further improve the efficiency of QML-AiNet, we also modify the mutation operator according to the features of discrete qualitative model space. Experimental results show that the performance of QML-AiNet is comparable to QML-CLONALG, a QML system using the clonal selection algorithm (CLONALG). More importantly, QML-AiNet with the modified mutation operator can significantly improve the scalability of QML and is much more efficient than QML-CLONALG.

Net Generation's Learning Styles in Nursing Education.

Science.gov (United States)

Christodoulou, Eleni; Kalokairinou, Athina

2015-01-01

Numerous surveys have confirmed that emerging technologies and Web 2.0 tools have been a defining feature in the lives of current students, estimating that there is a fundamental shift in the way young people communicate, socialize and learn. Nursing students in higher education are characterized as digital literate with distinct traits which influence their learning styles. Millennials exhibit distinct learning preferences such as teamwork, experiential activities, structure, instant feedback and technology integration. Higher education institutions should be aware of the implications of the Net Generation coming to university and be prepared to meet their expectations and learning needs.

Processing of chromatic information in a deep convolutional neural network.

Science.gov (United States)

Flachot, Alban; Gegenfurtner, Karl R

2018-04-01

Deep convolutional neural networks are a class of machine-learning algorithms capable of solving non-trivial tasks, such as object recognition, with human-like performance. Little is known about the exact computations that deep neural networks learn, and to what extent these computations are similar to the ones performed by the primate brain. Here, we investigate how color information is processed in the different layers of the AlexNet deep neural network, originally trained on object classification of over 1.2M images of objects in their natural contexts. We found that the color-responsive units in the first layer of AlexNet learned linear features and were broadly tuned to two directions in color space, analogously to what is known of color responsive cells in the primate thalamus. Moreover, these directions are decorrelated and lead to statistically efficient representations, similar to the cardinal directions of the second-stage color mechanisms in primates. We also found, in analogy to the early stages of the primate visual system, that chromatic and achromatic information were segregated in the early layers of the network. Units in the higher layers of AlexNet exhibit on average a lower responsivity for color than units at earlier stages.

Machine Learning Topological Invariants with Neural Networks

Science.gov (United States)

Zhang, Pengfei; Shen, Huitao; Zhai, Hui

2018-02-01

In this Letter we supervisedly train neural networks to distinguish different topological phases in the context of topological band insulators. After training with Hamiltonians of one-dimensional insulators with chiral symmetry, the neural network can predict their topological winding numbers with nearly 100% accuracy, even for Hamiltonians with larger winding numbers that are not included in the training data. These results show a remarkable success that the neural network can capture the global and nonlinear topological features of quantum phases from local inputs. By opening up the neural network, we confirm that the network does learn the discrete version of the winding number formula. We also make a couple of remarks regarding the role of the symmetry and the opposite effect of regularization techniques when applying machine learning to physical systems.

Neural Net Safety Monitor Design

Science.gov (United States)

Larson, Richard R.

2007-01-01

The National Aeronautics and Space Administration (NASA) at the Dryden Flight Research Center (DFRC) has been conducting flight-test research using an F-15 aircraft (figure 1). This aircraft has been specially modified to interface a neural net (NN) controller as part of a single-string Airborne Research Test System (ARTS) computer with the existing quad-redundant flight control system (FCC) shown in figure 2. The NN commands are passed to FCC channels 2 and 4 and are cross channel data linked (CCDL) to the other computers as shown. Numerous types of fault-detection monitors exist in the FCC when the NN mode is engaged; these monitors would cause an automatic disengagement of the NN in the event of a triggering fault. Unfortunately, these monitors still may not prevent a possible NN hard-over command from coming through to the control laws. Therefore, an additional and unique safety monitor was designed for a single-string source that allows authority at maximum actuator rates but protects the pilot and structural loads against excessive g-limits in the case of a NN hard-over command input. This additional monitor resides in the FCCs and is executed before the control laws are computed. This presentation describes a floating limiter (FL) concept1 that was developed and successfully test-flown for this program (figure 3). The FL computes the rate of change of the NN commands that are input to the FCC from the ARTS. A window is created with upper and lower boundaries, which is constantly floating and trying to stay centered as the NN command rates are changing. The limiter works by only allowing the window to move at a much slower rate than those of the NN commands. Anywhere within the window, however, full rates are allowed. If a rate persists in one direction, it will eventually hit the boundary and be rate-limited to the floating limiter rate. When this happens, a persistent counter begins and after a limit is reached, a NN disengage command is generated. The

Schema generation in recurrent neural nets for intercepting a moving target.

Science.gov (United States)

Fleischer, Andreas G

2010-06-01

The grasping of a moving object requires the development of a motor strategy to anticipate the trajectory of the target and to compute an optimal course of interception. During the performance of perception-action cycles, a preprogrammed prototypical movement trajectory, a motor schema, may highly reduce the control load. Subjects were asked to hit a target that was moving along a circular path by means of a cursor. Randomized initial target positions and velocities were detected in the periphery of the eyes, resulting in a saccade toward the target. Even when the target disappeared, the eyes followed the target's anticipated course. The Gestalt of the trajectories was dependent on target velocity. The prediction capability of the motor schema was investigated by varying the visibility range of cursor and target. Motor schemata were determined to be of limited precision, and therefore visual feedback was continuously required to intercept the moving target. To intercept a target, the motor schema caused the hand to aim ahead and to adapt to the target trajectory. The control of cursor velocity determined the point of interception. From a modeling point of view, a neural network was developed that allowed the implementation of a motor schema interacting with feedback control in an iterative manner. The neural net of the Wilson type consists of an excitation-diffusion layer allowing the generation of a moving bubble. This activation bubble runs down an eye-centered motor schema and causes a planar arm model to move toward the target. A bubble provides local integration and straightening of the trajectory during repetitive moves. The schema adapts to task demands by learning and serves as forward controller. On the basis of these model considerations the principal problem of embedding motor schemata in generalized control strategies is discussed.

Continual Learning through Evolvable Neural Turing Machines

DEFF Research Database (Denmark)

Lüders, Benno; Schläger, Mikkel; Risi, Sebastian

2016-01-01

Continual learning, i.e. the ability to sequentially learn tasks without catastrophic forgetting of previously learned ones, is an important open challenge in machine learning. In this paper we take a step in this direction by showing that the recently proposed Evolving Neural Turing Machine (ENTM...

Neural Behavior Chain Learning of Mobile Robot Actions

Directory of Open Access Journals (Sweden)

Lejla Banjanovic-Mehmedovic

2012-01-01

Full Text Available This paper presents a visual/motor behavior learning approach, based on neural networks. We propose Behavior Chain Model (BCM in order to create a way of behavior learning. Our behavior-based system evolution task is a mobile robot detecting a target and driving/acting towards it. First, the mapping relations between the image feature domain of the object and the robot action domain are derived. Second, a multilayer neural network for offline learning of the mapping relations is used. This learning structure through neural network training process represents a connection between the visual perceptions and motor sequence of actions in order to grip a target. Last, using behavior learning through a noticed action chain, we can predict mobile robot behavior for a variety of similar tasks in similar environment. Prediction results suggest that the methodology is adequate and could be recognized as an idea for designing different mobile robot behaviour assistance.

Logarithmic learning for generalized classifier neural network.

Science.gov (United States)

Ozyildirim, Buse Melis; Avci, Mutlu

2014-12-01

Generalized classifier neural network is introduced as an efficient classifier among the others. Unless the initial smoothing parameter value is close to the optimal one, generalized classifier neural network suffers from convergence problem and requires quite a long time to converge. In this work, to overcome this problem, a logarithmic learning approach is proposed. The proposed method uses logarithmic cost function instead of squared error. Minimization of this cost function reduces the number of iterations used for reaching the minima. The proposed method is tested on 15 different data sets and performance of logarithmic learning generalized classifier neural network is compared with that of standard one. Thanks to operation range of radial basis function included by generalized classifier neural network, proposed logarithmic approach and its derivative has continuous values. This makes it possible to adopt the advantage of logarithmic fast convergence by the proposed learning method. Due to fast convergence ability of logarithmic cost function, training time is maximally decreased to 99.2%. In addition to decrease in training time, classification performance may also be improved till 60%. According to the test results, while the proposed method provides a solution for time requirement problem of generalized classifier neural network, it may also improve the classification accuracy. The proposed method can be considered as an efficient way for reducing the time requirement problem of generalized classifier neural network. Copyright © 2014 Elsevier Ltd. All rights reserved.

Combining deep residual neural network features with supervised machine learning algorithms to classify diverse food image datasets.

Science.gov (United States)

McAllister, Patrick; Zheng, Huiru; Bond, Raymond; Moorhead, Anne

2018-04-01

Obesity is increasing worldwide and can cause many chronic conditions such as type-2 diabetes, heart disease, sleep apnea, and some cancers. Monitoring dietary intake through food logging is a key method to maintain a healthy lifestyle to prevent and manage obesity. Computer vision methods have been applied to food logging to automate image classification for monitoring dietary intake. In this work we applied pretrained ResNet-152 and GoogleNet convolutional neural networks (CNNs), initially trained using ImageNet Large Scale Visual Recognition Challenge (ILSVRC) dataset with MatConvNet package, to extract features from food image datasets; Food 5K, Food-11, RawFooT-DB, and Food-101. Deep features were extracted from CNNs and used to train machine learning classifiers including artificial neural network (ANN), support vector machine (SVM), Random Forest, and Naive Bayes. Results show that using ResNet-152 deep features with SVM with RBF kernel can accurately detect food items with 99.4% accuracy using Food-5K validation food image dataset and 98.8% with Food-5K evaluation dataset using ANN, SVM-RBF, and Random Forest classifiers. Trained with ResNet-152 features, ANN can achieve 91.34%, 99.28% when applied to Food-11 and RawFooT-DB food image datasets respectively and SVM with RBF kernel can achieve 64.98% with Food-101 image dataset. From this research it is clear that using deep CNN features can be used efficiently for diverse food item image classification. The work presented in this research shows that pretrained ResNet-152 features provide sufficient generalisation power when applied to a range of food image classification tasks. Copyright © 2018 Elsevier Ltd. All rights reserved.

Learning-induced neural plasticity of speech processing before birth.

Science.gov (United States)

Partanen, Eino; Kujala, Teija; Näätänen, Risto; Liitola, Auli; Sambeth, Anke; Huotilainen, Minna

2013-09-10

Learning, the foundation of adaptive and intelligent behavior, is based on plastic changes in neural assemblies, reflected by the modulation of electric brain responses. In infancy, auditory learning implicates the formation and strengthening of neural long-term memory traces, improving discrimination skills, in particular those forming the prerequisites for speech perception and understanding. Although previous behavioral observations show that newborns react differentially to unfamiliar sounds vs. familiar sound material that they were exposed to as fetuses, the neural basis of fetal learning has not thus far been investigated. Here we demonstrate direct neural correlates of human fetal learning of speech-like auditory stimuli. We presented variants of words to fetuses; unlike infants with no exposure to these stimuli, the exposed fetuses showed enhanced brain activity (mismatch responses) in response to pitch changes for the trained variants after birth. Furthermore, a significant correlation existed between the amount of prenatal exposure and brain activity, with greater activity being associated with a higher amount of prenatal speech exposure. Moreover, the learning effect was generalized to other types of similar speech sounds not included in the training material. Consequently, our results indicate neural commitment specifically tuned to the speech features heard before birth and their memory representations.

Embedding responses in spontaneous neural activity shaped through sequential learning.

Directory of Open Access Journals (Sweden)

Tomoki Kurikawa

Full Text Available Recent experimental measurements have demonstrated that spontaneous neural activity in the absence of explicit external stimuli has remarkable spatiotemporal structure. This spontaneous activity has also been shown to play a key role in the response to external stimuli. To better understand this role, we proposed a viewpoint, "memories-as-bifurcations," that differs from the traditional "memories-as-attractors" viewpoint. Memory recall from the memories-as-bifurcations viewpoint occurs when the spontaneous neural activity is changed to an appropriate output activity upon application of an input, known as a bifurcation in dynamical systems theory, wherein the input modifies the flow structure of the neural dynamics. Learning, then, is a process that helps create neural dynamical systems such that a target output pattern is generated as an attractor upon a given input. Based on this novel viewpoint, we introduce in this paper an associative memory model with a sequential learning process. Using a simple hebbian-type learning, the model is able to memorize a large number of input/output mappings. The neural dynamics shaped through the learning exhibit different bifurcations to make the requested targets stable upon an increase in the input, and the neural activity in the absence of input shows chaotic dynamics with occasional approaches to the memorized target patterns. These results suggest that these dynamics facilitate the bifurcations to each target attractor upon application of the corresponding input, which thus increases the capacity for learning. This theoretical finding about the behavior of the spontaneous neural activity is consistent with recent experimental observations in which the neural activity without stimuli wanders among patterns evoked by previously applied signals. In addition, the neural networks shaped by learning properly reflect the correlations of input and target-output patterns in a similar manner to those designed in

End-to-end unsupervised deformable image registration with a convolutional neural network

NARCIS (Netherlands)

de Vos, Bob D.; Berendsen, Floris; Viergever, Max A.; Staring, Marius; Išgum, Ivana

2017-01-01

In this work we propose a deep learning network for deformable image registration (DIRNet). The DIRNet consists of a convolutional neural network (ConvNet) regressor, a spatial transformer, and a resampler. The ConvNet analyzes a pair of fixed and moving images and outputs parameters for the spatial

Neural-net disruption predictor in JT-60U

International Nuclear Information System (INIS)

Yoshino, R.

2003-01-01

The prediction of major disruptions caused by the density limit, the plasma current ramp-down with high internal inductance l i , the low density locked mode and the β-limit has been investigated in JT-60U. The concept of 'stability level', newly proposed in this paper to predict the occurrence of a major disruption, is calculated from nine input parameters every 2 ms by the neural network and the start of a major disruption is predicted when the stability level decreases to a certain level, the 'alarm level'. The neural network is trained in two steps. It is first trained with 12 disruptive and six non-disruptive shots (total of 8011 data points). Second, the target output data for 12 disruptive shots are modified and the network is trained again with additional data points generated by the operator. The 'neural-net disruption predictor' obtained has been tested for 300 disruptive shots (128 945 data points) and 1008 non-disruptive shots (982 800 data points) selected from nine years of operation (1991-1999) of JT-60U. Major disruptions except for those caused by the -limit have been predicted with a prediction success rate of 97-98% at 10 ms prior to the disruption and higher than 90% at 30 ms prior to the disruption while the false alarm rate is 2.1% for non-disruptive shots. This prediction performance has been confirmed for 120 disruptive shots (56 163 data points), caused by the density limit, as well as 1032 non-disruptive shots (1004 611 data points) in the last four years of operation (1999-2002) of JT-60U. A careful selection of the input parameters supplied to the network and the newly developed two-step training of the network have reduced the false alarm rate resulting in a considerable improvement of the prediction success rate. (author)

Convolutional neural network with transfer learning for rice type classification

Science.gov (United States)

Patel, Vaibhav Amit; Joshi, Manjunath V.

2018-04-01

Presently, rice type is identified manually by humans, which is time consuming and error prone. Therefore, there is a need to do this by machine which makes it faster with greater accuracy. This paper proposes a deep learning based method for classification of rice types. We propose two methods to classify the rice types. In the first method, we train a deep convolutional neural network (CNN) using the given segmented rice images. In the second method, we train a combination of a pretrained VGG16 network and the proposed method, while using transfer learning in which the weights of a pretrained network are used to achieve better accuracy. Our approach can also be used for classification of rice grain as broken or fine. We train a 5-class model for classifying rice types using 4000 training images and another 2- class model for the classification of broken and normal rice using 1600 training images. We observe that despite having distinct rice images, our architecture, pretrained on ImageNet data boosts classification accuracy significantly.

Windowed active sampling for reliable neural learning

NARCIS (Netherlands)

Barakova, E.I; Spaanenburg, L

The composition of the example set has a major impact on the quality of neural learning. The popular approach is focused on extensive pre-processing to bridge the representation gap between process measurement and neural presentation. In contrast, windowed active sampling attempts to solve these

Goal-seeking neural net for recall and recognition

Science.gov (United States)

Omidvar, Omid M.

1990-07-01

Neural networks have been used to mimic cognitive processes which take place in animal brains. The learning capability inherent in neural networks makes them suitable candidates for adaptive tasks such as recall and recognition. The synaptic reinforcements create a proper condition for adaptation, which results in memorization, formation of perception, and higher order information processing activities. In this research a model of a goal seeking neural network is studied and the operation of the network with regard to recall and recognition is analyzed. In these analyses recall is defined as retrieval of stored information where little or no matching is involved. On the other hand recognition is recall with matching; therefore it involves memorizing a piece of information with complete presentation. This research takes the generalized view of reinforcement in which all the signals are potential reinforcers. The neuronal response is considered to be the source of the reinforcement. This local approach to adaptation leads to the goal seeking nature of the neurons as network components. In the proposed model all the synaptic strengths are reinforced in parallel while the reinforcement among the layers is done in a distributed fashion and pipeline mode from the last layer inward. A model of complex neuron with varying threshold is developed to account for inhibitory and excitatory behavior of real neuron. A goal seeking model of a neural network is presented. This network is utilized to perform recall and recognition tasks. The performance of the model with regard to the assigned tasks is presented.

AggNet: Deep Learning From Crowds for Mitosis Detection in Breast Cancer Histology Images.

Science.gov (United States)

Albarqouni, Shadi; Baur, Christoph; Achilles, Felix; Belagiannis, Vasileios; Demirci, Stefanie; Navab, Nassir

2016-05-01

The lack of publicly available ground-truth data has been identified as the major challenge for transferring recent developments in deep learning to the biomedical imaging domain. Though crowdsourcing has enabled annotation of large scale databases for real world images, its application for biomedical purposes requires a deeper understanding and hence, more precise definition of the actual annotation task. The fact that expert tasks are being outsourced to non-expert users may lead to noisy annotations introducing disagreement between users. Despite being a valuable resource for learning annotation models from crowdsourcing, conventional machine-learning methods may have difficulties dealing with noisy annotations during training. In this manuscript, we present a new concept for learning from crowds that handle data aggregation directly as part of the learning process of the convolutional neural network (CNN) via additional crowdsourcing layer (AggNet). Besides, we present an experimental study on learning from crowds designed to answer the following questions. 1) Can deep CNN be trained with data collected from crowdsourcing? 2) How to adapt the CNN to train on multiple types of annotation datasets (ground truth and crowd-based)? 3) How does the choice of annotation and aggregation affect the accuracy? Our experimental setup involved Annot8, a self-implemented web-platform based on Crowdflower API realizing image annotation tasks for a publicly available biomedical image database. Our results give valuable insights into the functionality of deep CNN learning from crowd annotations and prove the necessity of data aggregation integration.

SuperSpike: Supervised Learning in Multilayer Spiking Neural Networks.

Science.gov (United States)

Zenke, Friedemann; Ganguli, Surya

2018-04-13

A vast majority of computation in the brain is performed by spiking neural networks. Despite the ubiquity of such spiking, we currently lack an understanding of how biological spiking neural circuits learn and compute in vivo, as well as how we can instantiate such capabilities in artificial spiking circuits in silico. Here we revisit the problem of supervised learning in temporally coding multilayer spiking neural networks. First, by using a surrogate gradient approach, we derive SuperSpike, a nonlinear voltage-based three-factor learning rule capable of training multilayer networks of deterministic integrate-and-fire neurons to perform nonlinear computations on spatiotemporal spike patterns. Second, inspired by recent results on feedback alignment, we compare the performance of our learning rule under different credit assignment strategies for propagating output errors to hidden units. Specifically, we test uniform, symmetric, and random feedback, finding that simpler tasks can be solved with any type of feedback, while more complex tasks require symmetric feedback. In summary, our results open the door to obtaining a better scientific understanding of learning and computation in spiking neural networks by advancing our ability to train them to solve nonlinear problems involving transformations between different spatiotemporal spike time patterns.

Learning-parameter adjustment in neural networks

Science.gov (United States)

Heskes, Tom M.; Kappen, Bert

1992-06-01

We present a learning-parameter adjustment algorithm, valid for a large class of learning rules in neural-network literature. The algorithm follows directly from a consideration of the statistics of the weights in the network. The characteristic behavior of the algorithm is calculated, both in a fixed and a changing environment. A simple example, Widrow-Hoff learning for statistical classification, serves as an illustration.

Online neural monitoring of statistical learning.

Science.gov (United States)

Batterink, Laura J; Paller, Ken A

2017-05-01

The extraction of patterns in the environment plays a critical role in many types of human learning, from motor skills to language acquisition. This process is known as statistical learning. Here we propose that statistical learning has two dissociable components: (1) perceptual binding of individual stimulus units into integrated composites and (2) storing those integrated representations for later use. Statistical learning is typically assessed using post-learning tasks, such that the two components are conflated. Our goal was to characterize the online perceptual component of statistical learning. Participants were exposed to a structured stream of repeating trisyllabic nonsense words and a random syllable stream. Online learning was indexed by an EEG-based measure that quantified neural entrainment at the frequency of the repeating words relative to that of individual syllables. Statistical learning was subsequently assessed using conventional measures in an explicit rating task and a reaction-time task. In the structured stream, neural entrainment to trisyllabic words was higher than in the random stream, increased as a function of exposure to track the progression of learning, and predicted performance on the reaction time (RT) task. These results demonstrate that monitoring this critical component of learning via rhythmic EEG entrainment reveals a gradual acquisition of knowledge whereby novel stimulus sequences are transformed into familiar composites. This online perceptual transformation is a critical component of learning. Copyright © 2017 Elsevier Ltd. All rights reserved.

Prediction of Disease Causing Non-Synonymous SNPs by the Artificial Neural Network Predictor NetDiseaseSNP

DEFF Research Database (Denmark)

Johansen, Morten Bo; Gonzalez-Izarzugaza, Jose Maria; Brunak, Søren

2013-01-01

We have developed a sequence conservation-based artificial neural network predictor called NetDiseaseSNP which classifies nsSNPs as disease-causing or neutral. Our method uses the excellent alignment generation algorithm of SIFT to identify related sequences and a combination of 31 features...

Accelerator diagnosis and control by Neural Nets

International Nuclear Information System (INIS)

Spencer, J.E.

1989-01-01

Neural Nets (NN) have been described as a solution looking for a problem. In the last conference, Artificial Intelligence (AI) was considered in the accelerator context. While good for local surveillance and control, its use for large complex systems (LCS) was much more restricted. By contrast, NN provide a good metaphor for LCS. It can be argued that they are logically equivalent to multi-loop feedback/forward control of faulty systems, and therefore provide an ideal adaptive control system. Thus, where AI may be good for maintaining a 'golden orbit,' NN should be good for obtaining it via a quantitative approach to 'look and adjust' methods like operator tweaking which use pattern recognition to deal with hardware and software limitations, inaccuracies or errors as well as imprecise knowledge or understanding of effects like annealing and hysteresis. Further, insights from NN allow one to define feasibility conditions for LCS in terms of design constraints and tolerances. Hardware and software implications are discussed and several LCS of current interest are compared and contrasted. 15 refs., 5 figs

Accelerator diagnosis and control by Neural Nets

International Nuclear Information System (INIS)

Spencer, J.E.

1989-01-01

Neural Nets (NN) have been described as a solution looking for a problem. In the last conference, Artificial Intelligence (AI) was considered in the accelerator context. While good for local surveillance and control, its use for large complex systems (LCS) was much more restricted. By contrast, NN provide a good metaphore for LCS. It can be argued that they are logically equivalent to multi-loop feedback/forward control of faulty systems and therefore provide an ideal adaptive control system. Thus, where A1 may be good for maintaining a golden orbit, NN should be good for obtaining it via a quantitative approach to look and adjust methods like operator tweaking which use pattern recognition to deal with hardware and software limitations, inaccuracies or errors as well as imprecise knowledge or understanding of effects like annealing and hysteresis. Further, insights from NN allow one to define feasibility conditions for LCS in terms of design constraints and tolerances. Hardware and software implications are discussed and several LCS of current interest are compared and contrasted. 15 refs., 5 figs

Prediction of disease causing non-synonymous SNPs by the Artificial Neural Network Predictor NetDiseaseSNP.

Directory of Open Access Journals (Sweden)

Morten Bo Johansen

Full Text Available We have developed a sequence conservation-based artificial neural network predictor called NetDiseaseSNP which classifies nsSNPs as disease-causing or neutral. Our method uses the excellent alignment generation algorithm of SIFT to identify related sequences and a combination of 31 features assessing sequence conservation and the predicted surface accessibility to produce a single score which can be used to rank nsSNPs based on their potential to cause disease. NetDiseaseSNP classifies successfully disease-causing and neutral mutations. In addition, we show that NetDiseaseSNP discriminates cancer driver and passenger mutations satisfactorily. Our method outperforms other state-of-the-art methods on several disease/neutral datasets as well as on cancer driver/passenger mutation datasets and can thus be used to pinpoint and prioritize plausible disease candidates among nsSNPs for further investigation. NetDiseaseSNP is publicly available as an online tool as well as a web service: http://www.cbs.dtu.dk/services/NetDiseaseSNP.

Neural Networks for Self-tuning Control Systems

Directory of Open Access Journals (Sweden)

A. Noriega Ponce

2004-01-01

Full Text Available In this paper, we presented a self-tuning control algorithm based on a three layers perceptron type neural network. The proposed algorithm is advantageous in the sense that practically a previous training of the net is not required and some changes in the set-point are generally enough to adjust the learning coefficient. Optionally, it is possible to introduce a self-tuning mechanism of the learning coefficient although by the moment it is not possible to give final conclusions about this possibility. The proposed algorithm has the special feature that the regulation error instead of the net output error is retropropagated for the weighting coefficients modifications. 

Development switch in neural circuitry underlying odor-malaise learning.

Science.gov (United States)

Shionoya, Kiseko; Moriceau, Stephanie; Lunday, Lauren; Miner, Cathrine; Roth, Tania L; Sullivan, Regina M

2006-01-01

Fetal and infant rats can learn to avoid odors paired with illness before development of brain areas supporting this learning in adults, suggesting an alternate learning circuit. Here we begin to document the transition from the infant to adult neural circuit underlying odor-malaise avoidance learning using LiCl (0.3 M; 1% of body weight, ip) and a 30-min peppermint-odor exposure. Conditioning groups included: Paired odor-LiCl, Paired odor-LiCl-Nursing, LiCl, and odor-saline. Results showed that Paired LiCl-odor conditioning induced a learned odor aversion in postnatal day (PN) 7, 12, and 23 pups. Odor-LiCl Paired Nursing induced a learned odor preference in PN7 and PN12 pups but blocked learning in PN23 pups. 14C 2-deoxyglucose (2-DG) autoradiography indicated enhanced olfactory bulb activity in PN7 and PN12 pups with odor preference and avoidance learning. The odor aversion in weanling aged (PN23) pups resulted in enhanced amygdala activity in Paired odor-LiCl pups, but not if they were nursing. Thus, the neural circuit supporting malaise-induced aversions changes over development, indicating that similar infant and adult-learned behaviors may have distinct neural circuits.

Feature to prototype transition in neural networks

Science.gov (United States)

Krotov, Dmitry; Hopfield, John

Models of associative memory with higher order (higher than quadratic) interactions, and their relationship to neural networks used in deep learning are discussed. Associative memory is conventionally described by recurrent neural networks with dynamical convergence to stable points. Deep learning typically uses feedforward neural nets without dynamics. However, a simple duality relates these two different views when applied to problems of pattern classification. From the perspective of associative memory such models deserve attention because they make it possible to store a much larger number of memories, compared to the quadratic case. In the dual description, these models correspond to feedforward neural networks with one hidden layer and unusual activation functions transmitting the activities of the visible neurons to the hidden layer. These activation functions are rectified polynomials of a higher degree rather than the rectified linear functions used in deep learning. The network learns representations of the data in terms of features for rectified linear functions, but as the power in the activation function is increased there is a gradual shift to a prototype-based representation, the two extreme regimes of pattern recognition known in cognitive psychology. Simons Center for Systems Biology.

Extraction of Built-Up Areas Using Convolutional Neural Networks and Transfer Learning from SENTINEL-2 Satellite Images

Science.gov (United States)

Bramhe, V. S.; Ghosh, S. K.; Garg, P. K.

2018-04-01

With rapid globalization, the extent of built-up areas is continuously increasing. Extraction of features for classifying built-up areas that are more robust and abstract is a leading research topic from past many years. Although, various studies have been carried out where spatial information along with spectral features has been utilized to enhance the accuracy of classification. Still, these feature extraction techniques require a large number of user-specific parameters and generally application specific. On the other hand, recently introduced Deep Learning (DL) techniques requires less number of parameters to represent more abstract aspects of the data without any manual effort. Since, it is difficult to acquire high-resolution datasets for applications that require large scale monitoring of areas. Therefore, in this study Sentinel-2 image has been used for built-up areas extraction. In this work, pre-trained Convolutional Neural Networks (ConvNets) i.e. Inception v3 and VGGNet are employed for transfer learning. Since these networks are trained on generic images of ImageNet dataset which are having very different characteristics from satellite images. Therefore, weights of networks are fine-tuned using data derived from Sentinel-2 images. To compare the accuracies with existing shallow networks, two state of art classifiers i.e. Gaussian Support Vector Machine (SVM) and Back-Propagation Neural Network (BP-NN) are also implemented. Both SVM and BP-NN gives 84.31 % and 82.86 % overall accuracies respectively. Inception-v3 and VGGNet gives 89.43 % of overall accuracy using fine-tuned VGGNet and 92.10 % when using Inception-v3. The results indicate high accuracy of proposed fine-tuned ConvNets on a 4-channel Sentinel-2 dataset for built-up area extraction.

EXTRACTION OF BUILT-UP AREAS USING CONVOLUTIONAL NEURAL NETWORKS AND TRANSFER LEARNING FROM SENTINEL-2 SATELLITE IMAGES

Directory of Open Access Journals (Sweden)

V. S. Bramhe

2018-04-01

Full Text Available With rapid globalization, the extent of built-up areas is continuously increasing. Extraction of features for classifying built-up areas that are more robust and abstract is a leading research topic from past many years. Although, various studies have been carried out where spatial information along with spectral features has been utilized to enhance the accuracy of classification. Still, these feature extraction techniques require a large number of user-specific parameters and generally application specific. On the other hand, recently introduced Deep Learning (DL techniques requires less number of parameters to represent more abstract aspects of the data without any manual effort. Since, it is difficult to acquire high-resolution datasets for applications that require large scale monitoring of areas. Therefore, in this study Sentinel-2 image has been used for built-up areas extraction. In this work, pre-trained Convolutional Neural Networks (ConvNets i.e. Inception v3 and VGGNet are employed for transfer learning. Since these networks are trained on generic images of ImageNet dataset which are having very different characteristics from satellite images. Therefore, weights of networks are fine-tuned using data derived from Sentinel-2 images. To compare the accuracies with existing shallow networks, two state of art classifiers i.e. Gaussian Support Vector Machine (SVM and Back-Propagation Neural Network (BP-NN are also implemented. Both SVM and BP-NN gives 84.31 % and 82.86 % overall accuracies respectively. Inception-v3 and VGGNet gives 89.43 % of overall accuracy using fine-tuned VGGNet and 92.10 % when using Inception-v3. The results indicate high accuracy of proposed fine-tuned ConvNets on a 4-channel Sentinel-2 dataset for built-up area extraction.

Unlearning in feed-forward multi-nets

NARCIS (Netherlands)

Spaanenburg, L; Kurkova,; Steele, NC; Neruda, R; Karny, M

2001-01-01

Multi-nets promise an improved performance over monolithic neural networks by virtue of their distributed implementation. Modular neural networks are multi-nets based on an judicious assembly of functionally different parts. This can be viewed as again a monolithic network, but with more complex

Parameter diagnostics of phases and phase transition learning by neural networks

Science.gov (United States)

Suchsland, Philippe; Wessel, Stefan

2018-05-01

We present an analysis of neural network-based machine learning schemes for phases and phase transitions in theoretical condensed matter research, focusing on neural networks with a single hidden layer. Such shallow neural networks were previously found to be efficient in classifying phases and locating phase transitions of various basic model systems. In order to rationalize the emergence of the classification process and for identifying any underlying physical quantities, it is feasible to examine the weight matrices and the convolutional filter kernels that result from the learning process of such shallow networks. Furthermore, we demonstrate how the learning-by-confusing scheme can be used, in combination with a simple threshold-value classification method, to diagnose the learning parameters of neural networks. In particular, we study the classification process of both fully-connected and convolutional neural networks for the two-dimensional Ising model with extended domain wall configurations included in the low-temperature regime. Moreover, we consider the two-dimensional XY model and contrast the performance of the learning-by-confusing scheme and convolutional neural networks trained on bare spin configurations to the case of preprocessed samples with respect to vortex configurations. We discuss these findings in relation to similar recent investigations and possible further applications.

Neural modularity helps organisms evolve to learn new skills without forgetting old skills.

Science.gov (United States)

Ellefsen, Kai Olav; Mouret, Jean-Baptiste; Clune, Jeff

2015-04-01

A long-standing goal in artificial intelligence is creating agents that can learn a variety of different skills for different problems. In the artificial intelligence subfield of neural networks, a barrier to that goal is that when agents learn a new skill they typically do so by losing previously acquired skills, a problem called catastrophic forgetting. That occurs because, to learn the new task, neural learning algorithms change connections that encode previously acquired skills. How networks are organized critically affects their learning dynamics. In this paper, we test whether catastrophic forgetting can be reduced by evolving modular neural networks. Modularity intuitively should reduce learning interference between tasks by separating functionality into physically distinct modules in which learning can be selectively turned on or off. Modularity can further improve learning by having a reinforcement learning module separate from sensory processing modules, allowing learning to happen only in response to a positive or negative reward. In this paper, learning takes place via neuromodulation, which allows agents to selectively change the rate of learning for each neural connection based on environmental stimuli (e.g. to alter learning in specific locations based on the task at hand). To produce modularity, we evolve neural networks with a cost for neural connections. We show that this connection cost technique causes modularity, confirming a previous result, and that such sparsely connected, modular networks have higher overall performance because they learn new skills faster while retaining old skills more and because they have a separate reinforcement learning module. Our results suggest (1) that encouraging modularity in neural networks may help us overcome the long-standing barrier of networks that cannot learn new skills without forgetting old ones, and (2) that one benefit of the modularity ubiquitous in the brains of natural animals might be to

Neural modularity helps organisms evolve to learn new skills without forgetting old skills.

Directory of Open Access Journals (Sweden)

Kai Olav Ellefsen

2015-04-01

Full Text Available A long-standing goal in artificial intelligence is creating agents that can learn a variety of different skills for different problems. In the artificial intelligence subfield of neural networks, a barrier to that goal is that when agents learn a new skill they typically do so by losing previously acquired skills, a problem called catastrophic forgetting. That occurs because, to learn the new task, neural learning algorithms change connections that encode previously acquired skills. How networks are organized critically affects their learning dynamics. In this paper, we test whether catastrophic forgetting can be reduced by evolving modular neural networks. Modularity intuitively should reduce learning interference between tasks by separating functionality into physically distinct modules in which learning can be selectively turned on or off. Modularity can further improve learning by having a reinforcement learning module separate from sensory processing modules, allowing learning to happen only in response to a positive or negative reward. In this paper, learning takes place via neuromodulation, which allows agents to selectively change the rate of learning for each neural connection based on environmental stimuli (e.g. to alter learning in specific locations based on the task at hand. To produce modularity, we evolve neural networks with a cost for neural connections. We show that this connection cost technique causes modularity, confirming a previous result, and that such sparsely connected, modular networks have higher overall performance because they learn new skills faster while retaining old skills more and because they have a separate reinforcement learning module. Our results suggest (1 that encouraging modularity in neural networks may help us overcome the long-standing barrier of networks that cannot learn new skills without forgetting old ones, and (2 that one benefit of the modularity ubiquitous in the brains of natural animals

Neural Modularity Helps Organisms Evolve to Learn New Skills without Forgetting Old Skills

Science.gov (United States)

Ellefsen, Kai Olav; Mouret, Jean-Baptiste; Clune, Jeff

2015-01-01

A long-standing goal in artificial intelligence is creating agents that can learn a variety of different skills for different problems. In the artificial intelligence subfield of neural networks, a barrier to that goal is that when agents learn a new skill they typically do so by losing previously acquired skills, a problem called catastrophic forgetting. That occurs because, to learn the new task, neural learning algorithms change connections that encode previously acquired skills. How networks are organized critically affects their learning dynamics. In this paper, we test whether catastrophic forgetting can be reduced by evolving modular neural networks. Modularity intuitively should reduce learning interference between tasks by separating functionality into physically distinct modules in which learning can be selectively turned on or off. Modularity can further improve learning by having a reinforcement learning module separate from sensory processing modules, allowing learning to happen only in response to a positive or negative reward. In this paper, learning takes place via neuromodulation, which allows agents to selectively change the rate of learning for each neural connection based on environmental stimuli (e.g. to alter learning in specific locations based on the task at hand). To produce modularity, we evolve neural networks with a cost for neural connections. We show that this connection cost technique causes modularity, confirming a previous result, and that such sparsely connected, modular networks have higher overall performance because they learn new skills faster while retaining old skills more and because they have a separate reinforcement learning module. Our results suggest (1) that encouraging modularity in neural networks may help us overcome the long-standing barrier of networks that cannot learn new skills without forgetting old ones, and (2) that one benefit of the modularity ubiquitous in the brains of natural animals might be to

Vicarious neural processing of outcomes during observational learning.

Directory of Open Access Journals (Sweden)

Elisabetta Monfardini

Full Text Available Learning what behaviour is appropriate in a specific context by observing the actions of others and their outcomes is a key constituent of human cognition, because it saves time and energy and reduces exposure to potentially dangerous situations. Observational learning of associative rules relies on the ability to map the actions of others onto our own, process outcomes, and combine these sources of information. Here, we combined newly developed experimental tasks and functional magnetic resonance imaging (fMRI to investigate the neural mechanisms that govern such observational learning. Results show that the neural systems involved in individual trial-and-error learning and in action observation and execution both participate in observational learning. In addition, we identified brain areas that specifically activate for others' incorrect outcomes during learning in the posterior medial frontal cortex (pMFC, the anterior insula and the posterior superior temporal sulcus (pSTS.

Vicarious neural processing of outcomes during observational learning.

Science.gov (United States)

Monfardini, Elisabetta; Gazzola, Valeria; Boussaoud, Driss; Brovelli, Andrea; Keysers, Christian; Wicker, Bruno

2013-01-01

Learning what behaviour is appropriate in a specific context by observing the actions of others and their outcomes is a key constituent of human cognition, because it saves time and energy and reduces exposure to potentially dangerous situations. Observational learning of associative rules relies on the ability to map the actions of others onto our own, process outcomes, and combine these sources of information. Here, we combined newly developed experimental tasks and functional magnetic resonance imaging (fMRI) to investigate the neural mechanisms that govern such observational learning. Results show that the neural systems involved in individual trial-and-error learning and in action observation and execution both participate in observational learning. In addition, we identified brain areas that specifically activate for others' incorrect outcomes during learning in the posterior medial frontal cortex (pMFC), the anterior insula and the posterior superior temporal sulcus (pSTS).

Learning and adaptation: neural and behavioural mechanisms behind behaviour change

Science.gov (United States)

Lowe, Robert; Sandamirskaya, Yulia

2018-01-01

This special issue presents perspectives on learning and adaptation as they apply to a number of cognitive phenomena including pupil dilation in humans and attention in robots, natural language acquisition and production in embodied agents (robots), human-robot game play and social interaction, neural-dynamic modelling of active perception and neural-dynamic modelling of infant development in the Piagetian A-not-B task. The aim of the special issue, through its contributions, is to highlight some of the critical neural-dynamic and behavioural aspects of learning as it grounds adaptive responses in robotic- and neural-dynamic systems.

Framing U-Net via Deep Convolutional Framelets: Application to Sparse-View CT.

Science.gov (United States)

Han, Yoseob; Ye, Jong Chul

2018-06-01

X-ray computed tomography (CT) using sparse projection views is a recent approach to reduce the radiation dose. However, due to the insufficient projection views, an analytic reconstruction approach using the filtered back projection (FBP) produces severe streaking artifacts. Recently, deep learning approaches using large receptive field neural networks such as U-Net have demonstrated impressive performance for sparse-view CT reconstruction. However, theoretical justification is still lacking. Inspired by the recent theory of deep convolutional framelets, the main goal of this paper is, therefore, to reveal the limitation of U-Net and propose new multi-resolution deep learning schemes. In particular, we show that the alternative U-Net variants such as dual frame and tight frame U-Nets satisfy the so-called frame condition which makes them better for effective recovery of high frequency edges in sparse-view CT. Using extensive experiments with real patient data set, we demonstrate that the new network architectures provide better reconstruction performance.

Comparison between extreme learning machine and wavelet neural networks in data classification

Science.gov (United States)

Yahia, Siwar; Said, Salwa; Jemai, Olfa; Zaied, Mourad; Ben Amar, Chokri

2017-03-01

Extreme learning Machine is a well known learning algorithm in the field of machine learning. It's about a feed forward neural network with a single-hidden layer. It is an extremely fast learning algorithm with good generalization performance. In this paper, we aim to compare the Extreme learning Machine with wavelet neural networks, which is a very used algorithm. We have used six benchmark data sets to evaluate each technique. These datasets Including Wisconsin Breast Cancer, Glass Identification, Ionosphere, Pima Indians Diabetes, Wine Recognition and Iris Plant. Experimental results have shown that both extreme learning machine and wavelet neural networks have reached good results.

Boltzmann learning of parameters in cellular neural networks

DEFF Research Database (Denmark)

Hansen, Lars Kai

1992-01-01

The use of Bayesian methods to design cellular neural networks for signal processing tasks and the Boltzmann machine learning rule for parameter estimation is discussed. The learning rule can be used for models with hidden units, or for completely unsupervised learning. The latter is exemplified...

Neural mechanisms of reinforcement learning in unmedicated patients with major depressive disorder.

Science.gov (United States)

Rothkirch, Marcus; Tonn, Jonas; Köhler, Stephan; Sterzer, Philipp

2017-04-01

According to current concepts, major depressive disorder is strongly related to dysfunctional neural processing of motivational information, entailing impairments in reinforcement learning. While computational modelling can reveal the precise nature of neural learning signals, it has not been used to study learning-related neural dysfunctions in unmedicated patients with major depressive disorder so far. We thus aimed at comparing the neural coding of reward and punishment prediction errors, representing indicators of neural learning-related processes, between unmedicated patients with major depressive disorder and healthy participants. To this end, a group of unmedicated patients with major depressive disorder (n = 28) and a group of age- and sex-matched healthy control participants (n = 30) completed an instrumental learning task involving monetary gains and losses during functional magnetic resonance imaging. The two groups did not differ in their learning performance. Patients and control participants showed the same level of prediction error-related activity in the ventral striatum and the anterior insula. In contrast, neural coding of reward prediction errors in the medial orbitofrontal cortex was reduced in patients. Moreover, neural reward prediction error signals in the medial orbitofrontal cortex and ventral striatum showed negative correlations with anhedonia severity. Using a standard instrumental learning paradigm we found no evidence for an overall impairment of reinforcement learning in medication-free patients with major depressive disorder. Importantly, however, the attenuated neural coding of reward in the medial orbitofrontal cortex and the relation between anhedonia and reduced reward prediction error-signalling in the medial orbitofrontal cortex and ventral striatum likely reflect an impairment in experiencing pleasure from rewarding events as a key mechanism of anhedonia in major depressive disorder. © The Author (2017). Published by Oxford

Fastest learning in small-world neural networks

International Nuclear Information System (INIS)

Simard, D.; Nadeau, L.; Kroeger, H.

2005-01-01

We investigate supervised learning in neural networks. We consider a multi-layered feed-forward network with back propagation. We find that the network of small-world connectivity reduces the learning error and learning time when compared to the networks of regular or random connectivity. Our study has potential applications in the domain of data-mining, image processing, speech recognition, and pattern recognition

ENERGY-NET (Energy, Environment and Society Learning Network): Best Practices to Enhance Informal Geoscience Learning

Science.gov (United States)

Rossi, R.; Elliott, E. M.; Bain, D.; Crowley, K. J.; Steiner, M. A.; Divers, M. T.; Hopkins, K. G.; Giarratani, L.; Gilmore, M. E.

2014-12-01

While energy links all living and non-living systems, the integration of energy, the environment, and society is often not clearly represented in 9 - 12 classrooms and informal learning venues. However, objective public learning that integrates these components is essential for improving public environmental literacy. ENERGY-NET (Energy, Environment and Society Learning Network) is a National Science Foundation funded initiative that uses an Earth Systems Science framework to guide experimental learning for high school students and to improve public learning opportunities regarding the energy-environment-society nexus in a Museum setting. One of the primary objectives of the ENERGY-NET project is to develop a rich set of experimental learning activities that are presented as exhibits at the Carnegie Museum of Natural History in Pittsburgh, Pennsylvania (USA). Here we detail the evolution of the ENERGY-NET exhibit building process and the subsequent evolution of exhibit content over the past three years. While preliminary plans included the development of five "exploration stations" (i.e., traveling activity carts) per calendar year, the opportunity arose to create a single, larger topical exhibit per semester, which was assumed to have a greater impact on museum visitors. Evaluative assessments conducted to date reveal important practices to be incorporated into ongoing exhibit development: 1) Undergraduate mentors and teen exhibit developers should receive additional content training to allow richer exhibit materials. 2) The development process should be distributed over as long a time period as possible and emphasize iteration. This project can serve as a model for other collaborations between geoscience departments and museums. In particular, these practices may streamline development of public presentations and increase the effectiveness of experimental learning activities.

Learning in neural networks based on a generalized fluctuation theorem

Science.gov (United States)

Hayakawa, Takashi; Aoyagi, Toshio

2015-11-01

Information maximization has been investigated as a possible mechanism of learning governing the self-organization that occurs within the neural systems of animals. Within the general context of models of neural systems bidirectionally interacting with environments, however, the role of information maximization remains to be elucidated. For bidirectionally interacting physical systems, universal laws describing the fluctuation they exhibit and the information they possess have recently been discovered. These laws are termed fluctuation theorems. In the present study, we formulate a theory of learning in neural networks bidirectionally interacting with environments based on the principle of information maximization. Our formulation begins with the introduction of a generalized fluctuation theorem, employing an interpretation appropriate for the present application, which differs from the original thermodynamic interpretation. We analytically and numerically demonstrate that the learning mechanism presented in our theory allows neural networks to efficiently explore their environments and optimally encode information about them.

Evolving Neural Turing Machines for Reward-based Learning

DEFF Research Database (Denmark)

Greve, Rasmus Boll; Jacobsen, Emil Juul; Risi, Sebastian

2016-01-01

An unsolved problem in neuroevolution (NE) is to evolve artificial neural networks (ANN) that can store and use information to change their behavior online. While plastic neural networks have shown promise in this context, they have difficulties retaining information over longer periods of time...... version of the double T-Maze, a complex reinforcement-like learning problem. In the T-Maze learning task the agent uses the memory bank to display adaptive behavior that normally requires a plastic ANN, thereby suggesting a complementary and effective mechanism for adaptive behavior in NE....

The EB factory project. I. A fast, neural-net-based, general purpose light curve classifier optimized for eclipsing binaries

International Nuclear Information System (INIS)

Paegert, Martin; Stassun, Keivan G.; Burger, Dan M.

2014-01-01

We describe a new neural-net-based light curve classifier and provide it with documentation as a ready-to-use tool for the community. While optimized for identification and classification of eclipsing binary stars, the classifier is general purpose, and has been developed for speed in the context of upcoming massive surveys such as the Large Synoptic Survey Telescope. A challenge for classifiers in the context of neural-net training and massive data sets is to minimize the number of parameters required to describe each light curve. We show that a simple and fast geometric representation that encodes the overall light curve shape, together with a chi-square parameter to capture higher-order morphology information results in efficient yet robust light curve classification, especially for eclipsing binaries. Testing the classifier on the ASAS light curve database, we achieve a retrieval rate of 98% and a false-positive rate of 2% for eclipsing binaries. We achieve similarly high retrieval rates for most other periodic variable-star classes, including RR Lyrae, Mira, and delta Scuti. However, the classifier currently has difficulty discriminating between different sub-classes of eclipsing binaries, and suffers a relatively low (∼60%) retrieval rate for multi-mode delta Cepheid stars. We find that it is imperative to train the classifier's neural network with exemplars that include the full range of light curve quality to which the classifier will be expected to perform; the classifier performs well on noisy light curves only when trained with noisy exemplars. The classifier source code, ancillary programs, a trained neural net, and a guide for use, are provided.

Threat driven modeling framework using petri nets for e-learning system.

Science.gov (United States)

Khamparia, Aditya; Pandey, Babita

2016-01-01

Vulnerabilities at various levels are main cause of security risks in e-learning system. This paper presents a modified threat driven modeling framework, to identify the threats after risk assessment which requires mitigation and how to mitigate those threats. To model those threat mitigations aspects oriented stochastic petri nets are used. This paper included security metrics based on vulnerabilities present in e-learning system. The Common Vulnerability Scoring System designed to provide a normalized method for rating vulnerabilities which will be used as basis in metric definitions and calculations. A case study has been also proposed which shows the need and feasibility of using aspect oriented stochastic petri net models for threat modeling which improves reliability, consistency and robustness of the e-learning system.

Quantitative phase microscopy using deep neural networks

Science.gov (United States)

Li, Shuai; Sinha, Ayan; Lee, Justin; Barbastathis, George

2018-02-01

Deep learning has been proven to achieve ground-breaking accuracy in various tasks. In this paper, we implemented a deep neural network (DNN) to achieve phase retrieval in a wide-field microscope. Our DNN utilized the residual neural network (ResNet) architecture and was trained using the data generated by a phase SLM. The results showed that our DNN was able to reconstruct the profile of the phase target qualitatively. In the meantime, large error still existed, which indicated that our approach still need to be improved.

An adaptable Boolean net trainable to control a computing robot

International Nuclear Information System (INIS)

Lauria, F. E.; Prevete, R.; Milo, M.; Visco, S.

1999-01-01

We discuss a method to implement in a Boolean neural network a Hebbian rule so to obtain an adaptable universal control system. We start by presenting both the Boolean neural net and the Hebbian rule we have considered. Then we discuss, first, the problems arising when the latter is naively implemented in a Boolean neural net, second, the method consenting us to overcome them and the ensuing adaptable Boolean neural net paradigm. Next, we present the adaptable Boolean neural net as an intelligent control system, actually controlling a writing robot, and discuss how to train it in the execution of the elementary arithmetic operations on operands represented by numerals with an arbitrary number of digits

Finite time convergent learning law for continuous neural networks.

Science.gov (United States)

Chairez, Isaac

2014-02-01

This paper addresses the design of a discontinuous finite time convergent learning law for neural networks with continuous dynamics. The neural network was used here to obtain a non-parametric model for uncertain systems described by a set of ordinary differential equations. The source of uncertainties was the presence of some external perturbations and poor knowledge of the nonlinear function describing the system dynamics. A new adaptive algorithm based on discontinuous algorithms was used to adjust the weights of the neural network. The adaptive algorithm was derived by means of a non-standard Lyapunov function that is lower semi-continuous and differentiable in almost the whole space. A compensator term was included in the identifier to reject some specific perturbations using a nonlinear robust algorithm. Two numerical examples demonstrated the improvements achieved by the learning algorithm introduced in this paper compared to classical schemes with continuous learning methods. The first one dealt with a benchmark problem used in the paper to explain how the discontinuous learning law works. The second one used the methane production model to show the benefits in engineering applications of the learning law proposed in this paper. Copyright © 2013 Elsevier Ltd. All rights reserved.

Neural Monkey: An Open-source Tool for Sequence Learning

Directory of Open Access Journals (Sweden)

Helcl Jindřich

2017-04-01

Full Text Available In this paper, we announce the development of Neural Monkey – an open-source neural machine translation (NMT and general sequence-to-sequence learning system built over the TensorFlow machine learning library. The system provides a high-level API tailored for fast prototyping of complex architectures with multiple sequence encoders and decoders. Models’ overall architecture is specified in easy-to-read configuration files. The long-term goal of the Neural Monkey project is to create and maintain a growing collection of implementations of recently proposed components or methods, and therefore it is designed to be easily extensible. Trained models can be deployed either for batch data processing or as a web service. In the presented paper, we describe the design of the system and introduce the reader to running experiments using Neural Monkey.

Genetic learning in rule-based and neural systems

Science.gov (United States)

Smith, Robert E.

1993-01-01

The design of neural networks and fuzzy systems can involve complex, nonlinear, and ill-conditioned optimization problems. Often, traditional optimization schemes are inadequate or inapplicable for such tasks. Genetic Algorithms (GA's) are a class of optimization procedures whose mechanics are based on those of natural genetics. Mathematical arguments show how GAs bring substantial computational leverage to search problems, without requiring the mathematical characteristics often necessary for traditional optimization schemes (e.g., modality, continuity, availability of derivative information, etc.). GA's have proven effective in a variety of search tasks that arise in neural networks and fuzzy systems. This presentation begins by introducing the mechanism and theoretical underpinnings of GA's. GA's are then related to a class of rule-based machine learning systems called learning classifier systems (LCS's). An LCS implements a low-level production-system that uses a GA as its primary rule discovery mechanism. This presentation illustrates how, despite its rule-based framework, an LCS can be thought of as a competitive neural network. Neural network simulator code for an LCS is presented. In this context, the GA is doing more than optimizing and objective function. It is searching for an ecology of hidden nodes with limited connectivity. The GA attempts to evolve this ecology such that effective neural network performance results. The GA is particularly well adapted to this task, given its naturally-inspired basis. The LCS/neural network analogy extends itself to other, more traditional neural networks. Conclusions to the presentation discuss the implications of using GA's in ecological search problems that arise in neural and fuzzy systems.

ConvNetQuake: Convolutional Neural Network for Earthquake Detection and Location

Science.gov (United States)

Denolle, M.; Perol, T.; Gharbi, M.

2017-12-01

Over the last decades, the volume of seismic data has increased exponentially, creating a need for efficient algorithms to reliably detect and locate earthquakes. Today's most elaborate methods scan through the plethora of continuous seismic records, searching for repeating seismic signals. In this work, we leverage the recent advances in artificial intelligence and present ConvNetQuake, a highly scalable convolutional neural network for probabilistic earthquake detection and location from single stations. We apply our technique to study two years of induced seismicity in Oklahoma (USA). We detect 20 times more earthquakes than previously cataloged by the Oklahoma Geological Survey. Our algorithm detection performances are at least one order of magnitude faster than other established methods.

Exploring the spatio-temporal neural basis of face learning

Science.gov (United States)

Yang, Ying; Xu, Yang; Jew, Carol A.; Pyles, John A.; Kass, Robert E.; Tarr, Michael J.

2017-01-01

Humans are experts at face individuation. Although previous work has identified a network of face-sensitive regions and some of the temporal signatures of face processing, as yet, we do not have a clear understanding of how such face-sensitive regions support learning at different time points. To study the joint spatio-temporal neural basis of face learning, we trained subjects to categorize two groups of novel faces and recorded their neural responses using magnetoencephalography (MEG) throughout learning. A regression analysis of neural responses in face-sensitive regions against behavioral learning curves revealed significant correlations with learning in the majority of the face-sensitive regions in the face network, mostly between 150–250 ms, but also after 300 ms. However, the effect was smaller in nonventral regions (within the superior temporal areas and prefrontal cortex) than that in the ventral regions (within the inferior occipital gyri (IOG), midfusiform gyri (mFUS) and anterior temporal lobes). A multivariate discriminant analysis also revealed that IOG and mFUS, which showed strong correlation effects with learning, exhibited significant discriminability between the two face categories at different time points both between 150–250 ms and after 300 ms. In contrast, the nonventral face-sensitive regions, where correlation effects with learning were smaller, did exhibit some significant discriminability, but mainly after 300 ms. In sum, our findings indicate that early and recurring temporal components arising from ventral face-sensitive regions are critically involved in learning new faces. PMID:28570739

A modified backpropagation algorithm for training neural networks on data with error bars

International Nuclear Information System (INIS)

Gernoth, K.A.; Clark, J.W.

1994-08-01

A method is proposed for training multilayer feedforward neural networks on data contaminated with noise. Specifically, we consider the case that the artificial neural system is required to learn a physical mapping when the available values of the target variable are subject to experimental uncertainties, but are characterized by error bars. The proposed method, based on maximum likelihood criterion for parameter estimation, involves simple modifications of the on-line backpropagation learning algorithm. These include incorporation of the error-bar assignments in a pattern-specific learning rate, together with epochal updating of a new measure of model accuracy that replaces the usual mean-square error. The extended backpropagation algorithm is successfully tested on two problems relevant to the modelling of atomic-mass systematics by neural networks. Provided the underlying mapping is reasonably smooth, neural nets trained with the new procedure are able to learn the true function to a good approximation even in the presence of high levels of Gaussian noise. (author). 26 refs, 2 figs, 5 tabs

Spaced Learning Enhances Subsequent Recognition Memory by Reducing Neural Repetition Suppression

Science.gov (United States)

Xue, Gui; Mei, Leilei; Chen, Chuansheng; Lu, Zhong-Lin; Poldrack, Russell; Dong, Qi

2011-01-01

Spaced learning usually leads to better recognition memory as compared with massed learning, yet the underlying neural mechanisms remain elusive. One open question is whether the spacing effect is achieved by reducing neural repetition suppression. In this fMRI study, participants were scanned while intentionally memorizing 120 novel faces, half…

Development of the neural net technique for particle physics. Study of the e+e- → Z0 → γH reaction

International Nuclear Information System (INIS)

Guicheney, C.

1992-01-01

This study is concerned with the application of pattern recognition methods through neural networks to High Energy physics. Two methods, Hopfield nets and multilayer nets, are analyzed and shown to have high potential for (resp.) clusterization and classification. Hopfield nets are used for the recognition of jets occurring during the fragmentation process of the e + e - reaction. Multilayer nets are used for the whole reaction analysis. Impediments are pointed out. Associated background noise is also examined. Multilayer nets may enhance the signal to noise ratio when looking for an upper limit for the production of a Higgs boson in the expected canal, and allow for the specific study of the γ b anti b

Thermodynamic efficiency of learning a rule in neural networks

Science.gov (United States)

Goldt, Sebastian; Seifert, Udo

2017-11-01

Biological systems have to build models from their sensory input data that allow them to efficiently process previously unseen inputs. Here, we study a neural network learning a binary classification rule for these inputs from examples provided by a teacher. We analyse the ability of the network to apply the rule to new inputs, that is to generalise from past experience. Using stochastic thermodynamics, we show that the thermodynamic costs of the learning process provide an upper bound on the amount of information that the network is able to learn from its teacher for both batch and online learning. This allows us to introduce a thermodynamic efficiency of learning. We analytically compute the dynamics and the efficiency of a noisy neural network performing online learning in the thermodynamic limit. In particular, we analyse three popular learning algorithms, namely Hebbian, Perceptron and AdaTron learning. Our work extends the methods of stochastic thermodynamics to a new type of learning problem and might form a suitable basis for investigating the thermodynamics of decision-making.

Convolutional Neural Network for Histopathological Analysis of Osteosarcoma.

Science.gov (United States)

Mishra, Rashika; Daescu, Ovidiu; Leavey, Patrick; Rakheja, Dinesh; Sengupta, Anita

2018-03-01

Pathologists often deal with high complexity and sometimes disagreement over osteosarcoma tumor classification due to cellular heterogeneity in the dataset. Segmentation and classification of histology tissue in H&E stained tumor image datasets is a challenging task because of intra-class variations, inter-class similarity, crowded context, and noisy data. In recent years, deep learning approaches have led to encouraging results in breast cancer and prostate cancer analysis. In this article, we propose convolutional neural network (CNN) as a tool to improve efficiency and accuracy of osteosarcoma tumor classification into tumor classes (viable tumor, necrosis) versus nontumor. The proposed CNN architecture contains eight learned layers: three sets of stacked two convolutional layers interspersed with max pooling layers for feature extraction and two fully connected layers with data augmentation strategies to boost performance. The use of a neural network results in higher accuracy of average 92% for the classification. We compare the proposed architecture with three existing and proven CNN architectures for image classification: AlexNet, LeNet, and VGGNet. We also provide a pipeline to calculate percentage necrosis in a given whole slide image. We conclude that the use of neural networks can assure both high accuracy and efficiency in osteosarcoma classification.

Hebbian learning of hand-centred representations in a hierarchical neural network model of the primate visual system

Science.gov (United States)

Born, Jannis; Stringer, Simon M.

2017-01-01

A subset of neurons in the posterior parietal and premotor areas of the primate brain respond to the locations of visual targets in a hand-centred frame of reference. Such hand-centred visual representations are thought to play an important role in visually-guided reaching to target locations in space. In this paper we show how a biologically plausible, Hebbian learning mechanism may account for the development of localized hand-centred representations in a hierarchical neural network model of the primate visual system, VisNet. The hand-centered neurons developed in the model use an invariance learning mechanism known as continuous transformation (CT) learning. In contrast to previous theoretical proposals for the development of hand-centered visual representations, CT learning does not need a memory trace of recent neuronal activity to be incorporated in the synaptic learning rule. Instead, CT learning relies solely on a Hebbian learning rule, which is able to exploit the spatial overlap that naturally occurs between successive images of a hand-object configuration as it is shifted across different retinal locations due to saccades. Our simulations show how individual neurons in the network model can learn to respond selectively to target objects in particular locations with respect to the hand, irrespective of where the hand-object configuration occurs on the retina. The response properties of these hand-centred neurons further generalise to localised receptive fields in the hand-centred space when tested on novel hand-object configurations that have not been explored during training. Indeed, even when the network is trained with target objects presented across a near continuum of locations around the hand during training, the model continues to develop hand-centred neurons with localised receptive fields in hand-centred space. With the help of principal component analysis, we provide the first theoretical framework that explains the behavior of Hebbian learning

Hebbian learning of hand-centred representations in a hierarchical neural network model of the primate visual system.

Science.gov (United States)

Born, Jannis; Galeazzi, Juan M; Stringer, Simon M

2017-01-01

A subset of neurons in the posterior parietal and premotor areas of the primate brain respond to the locations of visual targets in a hand-centred frame of reference. Such hand-centred visual representations are thought to play an important role in visually-guided reaching to target locations in space. In this paper we show how a biologically plausible, Hebbian learning mechanism may account for the development of localized hand-centred representations in a hierarchical neural network model of the primate visual system, VisNet. The hand-centered neurons developed in the model use an invariance learning mechanism known as continuous transformation (CT) learning. In contrast to previous theoretical proposals for the development of hand-centered visual representations, CT learning does not need a memory trace of recent neuronal activity to be incorporated in the synaptic learning rule. Instead, CT learning relies solely on a Hebbian learning rule, which is able to exploit the spatial overlap that naturally occurs between successive images of a hand-object configuration as it is shifted across different retinal locations due to saccades. Our simulations show how individual neurons in the network model can learn to respond selectively to target objects in particular locations with respect to the hand, irrespective of where the hand-object configuration occurs on the retina. The response properties of these hand-centred neurons further generalise to localised receptive fields in the hand-centred space when tested on novel hand-object configurations that have not been explored during training. Indeed, even when the network is trained with target objects presented across a near continuum of locations around the hand during training, the model continues to develop hand-centred neurons with localised receptive fields in hand-centred space. With the help of principal component analysis, we provide the first theoretical framework that explains the behavior of Hebbian learning

Random neural Q-learning for obstacle avoidance of a mobile robot in unknown environments

Directory of Open Access Journals (Sweden)

Jing Yang

2016-07-01

Full Text Available The article presents a random neural Q-learning strategy for the obstacle avoidance problem of an autonomous mobile robot in unknown environments. In the proposed strategy, two independent modules, namely, avoidance without considering the target and goal-seeking without considering obstacles, are first trained using the proposed random neural Q-learning algorithm to obtain their best control policies. Then, the two trained modules are combined based on a switching function to realize the obstacle avoidance in unknown environments. For the proposed random neural Q-learning algorithm, a single-hidden layer feedforward network is used to approximate the Q-function to estimate the Q-value. The parameters of the single-hidden layer feedforward network are modified using the recently proposed neural algorithm named the online sequential version of extreme learning machine, where the parameters of the hidden nodes are assigned randomly and the sample data can come one by one. However, different from the original online sequential version of extreme learning machine algorithm, the initial output weights are estimated subjected to quadratic inequality constraint to improve the convergence speed. Finally, the simulation results demonstrate that the proposed random neural Q-learning strategy can successfully solve the obstacle avoidance problem. Also, the higher learning efficiency and better generalization ability are achieved by the proposed random neural Q-learning algorithm compared with the Q-learning based on the back-propagation method.

Artificial neural net system for interactive tissue classification with MR imaging and image segmentation

International Nuclear Information System (INIS)

Clarke, L.P.; Silbiger, M.; Naylor, C.; Brown, K.

1990-01-01

This paper reports on the development of interactive methods for MR tissue classification that permit mathematically rigorous methods for three-dimensional image segmentation and automatic organ/tumor contouring, as required for surgical and RTP planning. The authors investigate a number of image-intensity based tissue- classification methods that make no implicit assumptions on the MR parameters and hence are not limited by image data set. Similarly, we have trained artificial neural net (ANN) systems for both supervised and unsupervised tissue classification

Deep learning in neural networks: an overview.

Science.gov (United States)

Schmidhuber, Jürgen

2015-01-01

In recent years, deep artificial neural networks (including recurrent ones) have won numerous contests in pattern recognition and machine learning. This historical survey compactly summarizes relevant work, much of it from the previous millennium. Shallow and Deep Learners are distinguished by the depth of their credit assignment paths, which are chains of possibly learnable, causal links between actions and effects. I review deep supervised learning (also recapitulating the history of backpropagation), unsupervised learning, reinforcement learning & evolutionary computation, and indirect search for short programs encoding deep and large networks.

Evolutionary pruning of transfer learned deep convolutional neural network for breast cancer diagnosis in digital breast tomosynthesis.

Science.gov (United States)

Samala, Ravi K; Chan, Heang-Ping; Hadjiiski, Lubomir M; Helvie, Mark A; Richter, Caleb; Cha, Kenny

2018-05-01

Deep learning models are highly parameterized, resulting in difficulty in inference and transfer learning for image recognition tasks. In this work, we propose a layered pathway evolution method to compress a deep convolutional neural network (DCNN) for classification of masses in digital breast tomosynthesis (DBT). The objective is to prune the number of tunable parameters while preserving the classification accuracy. In the first stage transfer learning, 19 632 augmented regions-of-interest (ROIs) from 2454 mass lesions on mammograms were used to train a pre-trained DCNN on ImageNet. In the second stage transfer learning, the DCNN was used as a feature extractor followed by feature selection and random forest classification. The pathway evolution was performed using genetic algorithm in an iterative approach with tournament selection driven by count-preserving crossover and mutation. The second stage was trained with 9120 DBT ROIs from 228 mass lesions using leave-one-case-out cross-validation. The DCNN was reduced by 87% in the number of neurons, 34% in the number of parameters, and 95% in the number of multiply-and-add operations required in the convolutional layers. The test AUC on 89 mass lesions from 94 independent DBT cases before and after pruning were 0.88 and 0.90, respectively, and the difference was not statistically significant (p  >  0.05). The proposed DCNN compression approach can reduce the number of required operations by 95% while maintaining the classification performance. The approach can be extended to other deep neural networks and imaging tasks where transfer learning is appropriate.

Evolutionary pruning of transfer learned deep convolutional neural network for breast cancer diagnosis in digital breast tomosynthesis

Science.gov (United States)

Samala, Ravi K.; Chan, Heang-Ping; Hadjiiski, Lubomir M.; Helvie, Mark A.; Richter, Caleb; Cha, Kenny

2018-05-01

Deep learning models are highly parameterized, resulting in difficulty in inference and transfer learning for image recognition tasks. In this work, we propose a layered pathway evolution method to compress a deep convolutional neural network (DCNN) for classification of masses in digital breast tomosynthesis (DBT). The objective is to prune the number of tunable parameters while preserving the classification accuracy. In the first stage transfer learning, 19 632 augmented regions-of-interest (ROIs) from 2454 mass lesions on mammograms were used to train a pre-trained DCNN on ImageNet. In the second stage transfer learning, the DCNN was used as a feature extractor followed by feature selection and random forest classification. The pathway evolution was performed using genetic algorithm in an iterative approach with tournament selection driven by count-preserving crossover and mutation. The second stage was trained with 9120 DBT ROIs from 228 mass lesions using leave-one-case-out cross-validation. The DCNN was reduced by 87% in the number of neurons, 34% in the number of parameters, and 95% in the number of multiply-and-add operations required in the convolutional layers. The test AUC on 89 mass lesions from 94 independent DBT cases before and after pruning were 0.88 and 0.90, respectively, and the difference was not statistically significant (p  >  0.05). The proposed DCNN compression approach can reduce the number of required operations by 95% while maintaining the classification performance. The approach can be extended to other deep neural networks and imaging tasks where transfer learning is appropriate.

Door and cabinet recognition using convolutional neural nets and real-time method fusion for handle detection and grasping

DEFF Research Database (Denmark)

Maurin, Adrian Llopart; Ravn, Ole; Andersen, Nils Axel

2017-01-01

In this paper we present a new method that robustly identifies doors, cabinets and their respective handles, with special emphasis on extracting useful features from handles to be then manipulated. The novelty of this system relies on the combination of a Convolutional Neural Net (CNN), as a form...

LOGIC WITH EXCEPTION ON THE ALGEBRA OF FOURIER-DUAL OPERATIONS: NEURAL NET MECHANISM OF COGNITIVE DISSONANCE REDUCING

Directory of Open Access Journals (Sweden)

A. V. Pavlov

2014-01-01

Full Text Available A mechanism of cognitive dissonance reducing is demonstrated with approach for non-monotonic fuzzy-valued logics by Fourier-holography technique implementation developing. Cognitive dissonance occurs under perceiving of new information that contradicts to the existing subjective pattern of the outside world, represented by double Fourier-transform cascade with a hologram – neural layers interconnections matrix of inner information representation and logical conclusion. The hologram implements monotonic logic according to “General Modus Ponens” rule. New information is represented by a hologram of exclusion that implements interconnections of logical conclusion and exclusion for neural layers. The latter are linked by Fourier transform that determines duality of the algebra forming operations of conjunction and disjunction. Hologram of exclusion forms conclusion that is dual to the “General Modus Ponens” conclusion. It is shown, that trained for the main rule and exclusion system can be represented by two-layered neural network with separate interconnection matrixes for direct and inverse iterations. The network energy function is involved determining the cyclic dynamics character; dissipative factor causing convergence type of the dynamics is analyzed. Both “General Modus Ponens” and exclusion holograms recording conditions on the dynamics and convergence of the system are demonstrated. The system converges to a stable status, in which logical conclusion doesn’t depend on the inner information. Such kind of dynamics, leading to tolerance forming, is typical for ordinary kind of thinking, aimed at inner pattern of outside world stability. For scientific kind of thinking, aimed at adequacy of the inner pattern of the world, a mechanism is needed to stop the net relaxation; the mechanism has to be external relative to the model of logic. Computer simulation results for the learning conditions adequate to real holograms recording are

www.HistoNet2000.de – didaktisches Konzept und Nutzerakzeptanz eines e-learning-Programms / www.HistoNet2000.de – didactic concept and user-acceptance of an e-learning program

Directory of Open Access Journals (Sweden)

Abele, Harald

2007-12-01

Full Text Available HistoNet2000 is an Internet learning and training program for Microscopic Anatomy. This online-platform was programmed as a prototype also to be used for other medical topics where visual learning plays an important role. The 2006 remodelled second version of HistoNet2000 is databased and has a tool for editors. In times of high student numbers but tight resources such as staff, classrooms and learning material HistoNet2000 supports the quality of and has a lasting effect on traditional teachings like lectures, classes etc. Furthermore it meets the growing wish of the students of information society for using multi-media systems by offering a blended-learning supply. The modular based program enables a linked and interactive as well as self-scrutinizing learning of Histology. The crucial visual training is supported by providing a wide range of pictures that cannot be offered in a book due to the high costs. As a tutor function is included the student has the possibility to communicate with the lecturer any time via e-mail, an offer widely used. Development and user-acceptance were scientifically analysed. A survey in 1998 and 1999 about e-learning asked 505 students in their preliminary clinical studies about hardware equipment, their attitude and desires in using the New Media. Though the hardware equipment was rather moderate at these times compared to nowadays, a majority showed quite an interest in the use of e-learning programs as a supplement to the traditional teaching methods and support for private study. The evaluation of the user-acceptance by logfiles 2006 and 2007 showed that HistoNet2000 is a very widely used learning program: the average of the more than 20,000 visitors every six months looked at about 100 pages and called up a data of more than 1 MB. In 2007 the user-acceptance even climbed over 40% in some months.

22nd Italian Workshop on Neural Nets

CERN Document Server

Bassis, Simone; Esposito, Anna; Morabito, Francesco

2013-01-01

This volume collects a selection of contributions which has been presented at the 22nd Italian Workshop on Neural Networks, the yearly meeting of the Italian Society for Neural Networks (SIREN). The conference was held in Italy, Vietri sul Mare (Salerno), during May 17-19, 2012. The annual meeting of SIREN is sponsored by International Neural Network Society (INNS), European Neural Network Society (ENNS) and IEEE Computational Intelligence Society (CIS). The book – as well as the workshop-  is organized in three main components, two special sessions and a group of regular sessions featuring different aspects and point of views of artificial neural networks and natural intelligence, also including applications of present compelling interest.

Maximum entropy methods for extracting the learned features of deep neural networks.

Science.gov (United States)

Finnegan, Alex; Song, Jun S

2017-10-01

New architectures of multilayer artificial neural networks and new methods for training them are rapidly revolutionizing the application of machine learning in diverse fields, including business, social science, physical sciences, and biology. Interpreting deep neural networks, however, currently remains elusive, and a critical challenge lies in understanding which meaningful features a network is actually learning. We present a general method for interpreting deep neural networks and extracting network-learned features from input data. We describe our algorithm in the context of biological sequence analysis. Our approach, based on ideas from statistical physics, samples from the maximum entropy distribution over possible sequences, anchored at an input sequence and subject to constraints implied by the empirical function learned by a network. Using our framework, we demonstrate that local transcription factor binding motifs can be identified from a network trained on ChIP-seq data and that nucleosome positioning signals are indeed learned by a network trained on chemical cleavage nucleosome maps. Imposing a further constraint on the maximum entropy distribution also allows us to probe whether a network is learning global sequence features, such as the high GC content in nucleosome-rich regions. This work thus provides valuable mathematical tools for interpreting and extracting learned features from feed-forward neural networks.

A Restricted Boltzman Neural Net to Infer Carbon Uptake from OCO-2 Satellite Data

Science.gov (United States)

Halem, M.; Dorband, J. E.; Radov, A.; Barr-Dallas, M.; Gentine, P.

2015-12-01

For several decades, scientists have been using satellite observations to infer climate budgets of terrestrial carbon uptake employing inverse methods in conjunction with ecosystem models and coupled global climate models. This is an extremely important Big Data calculation today since the net annual photosynthetic carbon uptake changes annually over land and removes on average ~20% of the emissions from human contributions to atmospheric loading of CO2 from fossil fuels. Unfortunately, such calculations have large uncertainties validated with in-situ networks of measuring stations across the globe. One difficulty in using satellite data for these budget calculations is that the models need to assimilate surface fluxes of CO2 as well as soil moisture, vegatation cover and the eddy covariance of latent and sensible heat to calculate the carbon fixed in the soil while satellite spectral observations only provide near surface concentrations of CO2. In July 2014, NASA successfully launched OCO-2 which provides 3km surface measurements of CO2 over land and oceans. We have collected nearly one year of Level 2 XCO2 data from the OCO-2 satellite for 3 sites of ~200 km2 at equatorial, temperate and high latitudes. Each selected site was part of the Fluxnet or ARM system with tower stations for measuring and collecting CO2 fluxes on an hourly basis, in addition to eddy transports of the other parameters. We are also planning to acquire the 4km NDVI products from MODIS and registering the data to the 3km XCO2 footprints for the three sites. We have implemented a restricted Boltzman machine on the quantum annealing D-Wave computer, a novel deep learning neural net, to be used for training with station data to infer CO2 fluxes from collocated XCO2, MODIS vegetative land cover and MERRA reanalysis surface exchange products. We will present performance assessments of the D-Wave Boltzman machine for generating XCO2 fluxes from the OCO-2 satellite observations for the 3 sites by

Learning drifting concepts with neural networks

NARCIS (Netherlands)

Biehl, Michael; Schwarze, Holm

1993-01-01

The learning of time-dependent concepts with a neural network is studied analytically and numerically. The linearly separable target rule is represented by an N-vector, whose time dependence is modelled by a random or deterministic drift process. A single-layer network is trained online using

Temporal-pattern learning in neural models

CERN Document Server

Genís, Carme Torras

1985-01-01

While the ability of animals to learn rhythms is an unquestionable fact, the underlying neurophysiological mechanisms are still no more than conjectures. This monograph explores the requirements of such mechanisms, reviews those previously proposed and postulates a new one based on a direct electric coding of stimulation frequencies. Experi­ mental support for the option taken is provided both at the single neuron and neural network levels. More specifically, the material presented divides naturally into four parts: a description of the experimental and theoretical framework where this work becomes meaningful (Chapter 2), a detailed specifica­ tion of the pacemaker neuron model proposed together with its valida­ tion through simulation (Chapter 3), an analytic study of the behavior of this model when submitted to rhythmic stimulation (Chapter 4) and a description of the neural network model proposed for learning, together with an analysis of the simulation results obtained when varying seve­ ral factors r...

Computational modeling of spiking neural network with learning rules from STDP and intrinsic plasticity

Science.gov (United States)

Li, Xiumin; Wang, Wei; Xue, Fangzheng; Song, Yongduan

2018-02-01

Recently there has been continuously increasing interest in building up computational models of spiking neural networks (SNN), such as the Liquid State Machine (LSM). The biologically inspired self-organized neural networks with neural plasticity can enhance the capability of computational performance, with the characteristic features of dynamical memory and recurrent connection cycles which distinguish them from the more widely used feedforward neural networks. Despite a variety of computational models for brain-like learning and information processing have been proposed, the modeling of self-organized neural networks with multi-neural plasticity is still an important open challenge. The main difficulties lie in the interplay among different forms of neural plasticity rules and understanding how structures and dynamics of neural networks shape the computational performance. In this paper, we propose a novel approach to develop the models of LSM with a biologically inspired self-organizing network based on two neural plasticity learning rules. The connectivity among excitatory neurons is adapted by spike-timing-dependent plasticity (STDP) learning; meanwhile, the degrees of neuronal excitability are regulated to maintain a moderate average activity level by another learning rule: intrinsic plasticity (IP). Our study shows that LSM with STDP+IP performs better than LSM with a random SNN or SNN obtained by STDP alone. The noticeable improvement with the proposed method is due to the better reflected competition among different neurons in the developed SNN model, as well as the more effectively encoded and processed relevant dynamic information with its learning and self-organizing mechanism. This result gives insights to the optimization of computational models of spiking neural networks with neural plasticity.

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.

Learning language with the wrong neural scaffolding: The cost of neural commitment to sounds.

Directory of Open Access Journals (Sweden)

Amy Sue Finn

2013-11-01

Full Text Available Does tuning to one’s native language explain the sensitive period for language learning? We explore the idea that tuning to (or becoming more selective for the properties of one’s native-language could result in being less open (or plastic for tuning to the properties of a new language. To explore how this might lead to the sensitive period for grammar learning, we ask if tuning to an earlier-learned aspect of language (sound structure has an impact on the neural representation of a later-learned aspect (grammar. English-speaking adults learned one of two miniature artificial languages over 4 days in the lab. Compared to English, both languages had novel grammar, but only one was comprised of novel sounds. After learning a language, participants were scanned while judging the grammaticality of sentences. Judgments were performed for the newly learned language and English. Learners of the similar-sounds language recruited regions that overlapped more with English. Learners of the distinct-sounds language, however, recruited the Superior Temporal Gyrus (STG to a greater extent, which was coactive with the Inferior Frontal Gyrus (IFG. Across learners, recruitment of IFG (but not STG predicted both learning success in tests conducted prior to the scan and grammatical judgment ability during the scan. Data suggest that adults’ difficulty learning language, especially grammar, could be due, at least in part, to the neural commitments they have made to the lower level linguistic components of their native language.

Learning language with the wrong neural scaffolding: the cost of neural commitment to sounds

Science.gov (United States)

Finn, Amy S.; Hudson Kam, Carla L.; Ettlinger, Marc; Vytlacil, Jason; D'Esposito, Mark

2013-01-01

Does tuning to one's native language explain the “sensitive period” for language learning? We explore the idea that tuning to (or becoming more selective for) the properties of one's native-language could result in being less open (or plastic) for tuning to the properties of a new language. To explore how this might lead to the sensitive period for grammar learning, we ask if tuning to an earlier-learned aspect of language (sound structure) has an impact on the neural representation of a later-learned aspect (grammar). English-speaking adults learned one of two miniature artificial languages (MALs) over 4 days in the lab. Compared to English, both languages had novel grammar, but only one was comprised of novel sounds. After learning a language, participants were scanned while judging the grammaticality of sentences. Judgments were performed for the newly learned language and English. Learners of the similar-sounds language recruited regions that overlapped more with English. Learners of the distinct-sounds language, however, recruited the Superior Temporal Gyrus (STG) to a greater extent, which was coactive with the Inferior Frontal Gyrus (IFG). Across learners, recruitment of IFG (but not STG) predicted both learning success in tests conducted prior to the scan and grammatical judgment ability during the scan. Data suggest that adults' difficulty learning language, especially grammar, could be due, at least in part, to the neural commitments they have made to the lower level linguistic components of their native language. PMID:24273497

Deep Learning Neural Networks in Cybersecurity - Managing Malware with AI

OpenAIRE

Rayle, Keith

2017-01-01

There’s a lot of talk about the benefits of deep learning (neural networks) and how it’s the new electricity that will power us into the future. Medical diagnosis, computer vision and speech recognition are all examples of use-cases where neural networks are being applied in our everyday business environment. This begs the question…what are the uses of neural-network applications for cyber security? How does the AI process work when applying neural networks to detect malicious software bombar...

Petri neural network model for the effect of controlled thermomechanical process parameters on the mechanical properties of HSLA steels

Energy Technology Data Exchange (ETDEWEB)

Datta, S.

1999-10-01

The effect of composition and controlled thermomechanical process parameters on the mechanical properties of HSLA steels is modelled using the Widrow-Hoff's concept of training a neural net with feed-forward topology by applying Rumelhart's back propagation type algorithm for supervised learning, using a Petri like net structure. The data used are from laboratory experiments as well as from the published literature. The results from the neural network are found to be consistent and in good agreement with the experimented results. (author)

Consensus-based distributed cooperative learning from closed-loop neural control systems.

Science.gov (United States)

Chen, Weisheng; Hua, Shaoyong; Zhang, Huaguang

2015-02-01

In this paper, the neural tracking problem is addressed for a group of uncertain nonlinear systems where the system structures are identical but the reference signals are different. This paper focuses on studying the learning capability of neural networks (NNs) during the control process. First, we propose a novel control scheme called distributed cooperative learning (DCL) control scheme, by establishing the communication topology among adaptive laws of NN weights to share their learned knowledge online. It is further proved that if the communication topology is undirected and connected, all estimated weights of NNs can converge to small neighborhoods around their optimal values over a domain consisting of the union of all state orbits. Second, as a corollary it is shown that the conclusion on the deterministic learning still holds in the decentralized adaptive neural control scheme where, however, the estimated weights of NNs just converge to small neighborhoods of the optimal values along their own state orbits. Thus, the learned controllers obtained by DCL scheme have the better generalization capability than ones obtained by decentralized learning method. A simulation example is provided to verify the effectiveness and advantages of the control schemes proposed in this paper.

Jet flavor tagging with Deep Learning using Python

CERN Multimedia

CERN. Geneva

2016-01-01

Besides the part that implements the resulting deep neural net in the ATLAS C++ software framework, a Python framework has been developed to connect HEP data to standard Data Science Python based libraries for Machine Learning. It makes use of HDF5, JSON and Pickle as intermediate data storage format, pandas and numpy for data handling and calculations, Keras for neural net construction and training as well as testing and matplotlib for plotting. It can be seen as an example of taking advantage of outside-HEP software developments without relying on the HEP standard ROOT.

Neural-Fitted TD-Leaf Learning for Playing Othello With Structured Neural Networks

NARCIS (Netherlands)

van den Dries, Sjoerd; Wiering, Marco A.

This paper describes a methodology for quickly learning to play games at a strong level. The methodology consists of a novel combination of three techniques, and a variety of experiments on the game of Othello demonstrates their usefulness. First, structures or topologies in neural network

Single-hidden-layer feed-forward quantum neural network based on Grover learning.

Science.gov (United States)

Liu, Cheng-Yi; Chen, Chein; Chang, Ching-Ter; Shih, Lun-Min

2013-09-01

In this paper, a novel single-hidden-layer feed-forward quantum neural network model is proposed based on some concepts and principles in the quantum theory. By combining the quantum mechanism with the feed-forward neural network, we defined quantum hidden neurons and connected quantum weights, and used them as the fundamental information processing unit in a single-hidden-layer feed-forward neural network. The quantum neurons make a wide range of nonlinear functions serve as the activation functions in the hidden layer of the network, and the Grover searching algorithm outstands the optimal parameter setting iteratively and thus makes very efficient neural network learning possible. The quantum neuron and weights, along with a Grover searching algorithm based learning, result in a novel and efficient neural network characteristic of reduced network, high efficient training and prospect application in future. Some simulations are taken to investigate the performance of the proposed quantum network and the result show that it can achieve accurate learning. Copyright © 2013 Elsevier Ltd. All rights reserved.

Learning sequential control in a Neural Blackboard Architecture for in situ concept reasoning

NARCIS (Netherlands)

van der Velde, Frank; van der Velde, Frank; Besold, Tarek R.; Lamb, Luis; Serafini, Luciano; Tabor, Whitney

2016-01-01

Simulations are presented and discussed of learning sequential control in a Neural Blackboard Architecture (NBA) for in situ concept-based reasoning. Sequential control is learned in a reservoir network, consisting of columns with neural circuits. This allows the reservoir to control the dynamics of

Teaching Russian Via Distance Learning, the EdNet Experience.

Science.gov (United States)

Zsiray, Stephen W., Jr.; And Others

In Utah, the statewide distance education network (EdNet) enables students from five rural and suburban high schools to learn Russian and earn college credits. Courses in Russian are offered through a partnership involving the Cache County School District, Utah State University, and the Utah State Office of Education. Classes are taught on one…

Learning-induced pattern classification in a chaotic neural network

International Nuclear Information System (INIS)

Li, Yang; Zhu, Ping; Xie, Xiaoping; He, Guoguang; Aihara, Kazuyuki

2012-01-01

In this Letter, we propose a Hebbian learning rule with passive forgetting (HLRPF) for use in a chaotic neural network (CNN). We then define the indices based on the Euclidean distance to investigate the evolution of the weights in a simplified way. Numerical simulations demonstrate that, under suitable external stimulations, the CNN with the proposed HLRPF acts as a fuzzy-like pattern classifier that performs much better than an ordinary CNN. The results imply relationship between learning and recognition. -- Highlights: ► Proposing a Hebbian learning rule with passive forgetting (HLRPF). ► Defining indices to investigate the evolution of the weights simply. ► The chaotic neural network with HLRPF acts as a fuzzy-like pattern classifier. ► The pattern classifier ability of the network is improved much.

Application of a neural network for reflectance spectrum classification

Science.gov (United States)

Yang, Gefei; Gartley, Michael

2017-05-01

Traditional reflectance spectrum classification algorithms are based on comparing spectrum across the electromagnetic spectrum anywhere from the ultra-violet to the thermal infrared regions. These methods analyze reflectance on a pixel by pixel basis. Inspired by high performance that Convolution Neural Networks (CNN) have demonstrated in image classification, we applied a neural network to analyze directional reflectance pattern images. By using the bidirectional reflectance distribution function (BRDF) data, we can reformulate the 4-dimensional into 2 dimensions, namely incident direction × reflected direction × channels. Meanwhile, RIT's micro-DIRSIG model is utilized to simulate additional training samples for improving the robustness of the neural networks training. Unlike traditional classification by using hand-designed feature extraction with a trainable classifier, neural networks create several layers to learn a feature hierarchy from pixels to classifier and all layers are trained jointly. Hence, the our approach of utilizing the angular features are different to traditional methods utilizing spatial features. Although training processing typically has a large computational cost, simple classifiers work well when subsequently using neural network generated features. Currently, most popular neural networks such as VGG, GoogLeNet and AlexNet are trained based on RGB spatial image data. Our approach aims to build a directional reflectance spectrum based neural network to help us to understand from another perspective. At the end of this paper, we compare the difference among several classifiers and analyze the trade-off among neural networks parameters.

Auto-Context Convolutional Neural Network (Auto-Net) for Brain Extraction in Magnetic Resonance Imaging.

Science.gov (United States)

Mohseni Salehi, Seyed Sadegh; Erdogmus, Deniz; Gholipour, Ali

2017-11-01

Brain extraction or whole brain segmentation is an important first step in many of the neuroimage analysis pipelines. The accuracy and the robustness of brain extraction, therefore, are crucial for the accuracy of the entire brain analysis process. The state-of-the-art brain extraction techniques rely heavily on the accuracy of alignment or registration between brain atlases and query brain anatomy, and/or make assumptions about the image geometry, and therefore have limited success when these assumptions do not hold or image registration fails. With the aim of designing an accurate, learning-based, geometry-independent, and registration-free brain extraction tool, in this paper, we present a technique based on an auto-context convolutional neural network (CNN), in which intrinsic local and global image features are learned through 2-D patches of different window sizes. We consider two different architectures: 1) a voxelwise approach based on three parallel 2-D convolutional pathways for three different directions (axial, coronal, and sagittal) that implicitly learn 3-D image information without the need for computationally expensive 3-D convolutions and 2) a fully convolutional network based on the U-net architecture. Posterior probability maps generated by the networks are used iteratively as context information along with the original image patches to learn the local shape and connectedness of the brain to extract it from non-brain tissue. The brain extraction results we have obtained from our CNNs are superior to the recently reported results in the literature on two publicly available benchmark data sets, namely, LPBA40 and OASIS, in which we obtained the Dice overlap coefficients of 97.73% and 97.62%, respectively. Significant improvement was achieved via our auto-context algorithm. Furthermore, we evaluated the performance of our algorithm in the challenging problem of extracting arbitrarily oriented fetal brains in reconstructed fetal brain magnetic

Bio-Inspired Neural Model for Learning Dynamic Models

Science.gov (United States)

Duong, Tuan; Duong, Vu; Suri, Ronald

2009-01-01

A neural-network mathematical model that, relative to prior such models, places greater emphasis on some of the temporal aspects of real neural physical processes, has been proposed as a basis for massively parallel, distributed algorithms that learn dynamic models of possibly complex external processes by means of learning rules that are local in space and time. The algorithms could be made to perform such functions as recognition and prediction of words in speech and of objects depicted in video images. The approach embodied in this model is said to be "hardware-friendly" in the following sense: The algorithms would be amenable to execution by special-purpose computers implemented as very-large-scale integrated (VLSI) circuits that would operate at relatively high speeds and low power demands.

Forecasting crude oil price with an EMD-based neural network ensemble learning paradigm

International Nuclear Information System (INIS)

Yu, Lean; Wang, Shouyang; Lai, Kin Keung

2008-01-01

In this study, an empirical mode decomposition (EMD) based neural network ensemble learning paradigm is proposed for world crude oil spot price forecasting. For this purpose, the original crude oil spot price series were first decomposed into a finite, and often small, number of intrinsic mode functions (IMFs). Then a three-layer feed-forward neural network (FNN) model was used to model each of the extracted IMFs, so that the tendencies of these IMFs could be accurately predicted. Finally, the prediction results of all IMFs are combined with an adaptive linear neural network (ALNN), to formulate an ensemble output for the original crude oil price series. For verification and testing, two main crude oil price series, West Texas Intermediate (WTI) crude oil spot price and Brent crude oil spot price, are used to test the effectiveness of the proposed EMD-based neural network ensemble learning methodology. Empirical results obtained demonstrate attractiveness of the proposed EMD-based neural network ensemble learning paradigm. (author)

Hebbian learning of hand-centred representations in a hierarchical neural network model of the primate visual system.

Directory of Open Access Journals (Sweden)

Jannis Born

Full Text Available A subset of neurons in the posterior parietal and premotor areas of the primate brain respond to the locations of visual targets in a hand-centred frame of reference. Such hand-centred visual representations are thought to play an important role in visually-guided reaching to target locations in space. In this paper we show how a biologically plausible, Hebbian learning mechanism may account for the development of localized hand-centred representations in a hierarchical neural network model of the primate visual system, VisNet. The hand-centered neurons developed in the model use an invariance learning mechanism known as continuous transformation (CT learning. In contrast to previous theoretical proposals for the development of hand-centered visual representations, CT learning does not need a memory trace of recent neuronal activity to be incorporated in the synaptic learning rule. Instead, CT learning relies solely on a Hebbian learning rule, which is able to exploit the spatial overlap that naturally occurs between successive images of a hand-object configuration as it is shifted across different retinal locations due to saccades. Our simulations show how individual neurons in the network model can learn to respond selectively to target objects in particular locations with respect to the hand, irrespective of where the hand-object configuration occurs on the retina. The response properties of these hand-centred neurons further generalise to localised receptive fields in the hand-centred space when tested on novel hand-object configurations that have not been explored during training. Indeed, even when the network is trained with target objects presented across a near continuum of locations around the hand during training, the model continues to develop hand-centred neurons with localised receptive fields in hand-centred space. With the help of principal component analysis, we provide the first theoretical framework that explains the behavior

Learning and coding in biological neural networks

Science.gov (United States)

Fiete, Ila Rani

How can large groups of neurons that locally modify their activities learn to collectively perform a desired task? Do studies of learning in small networks tell us anything about learning in the fantastically large collection of neurons that make up a vertebrate brain? What factors do neurons optimize by encoding sensory inputs or motor commands in the way they do? In this thesis I present a collection of four theoretical works: each of the projects was motivated by specific constraints and complexities of biological neural networks, as revealed by experimental studies; together, they aim to partially address some of the central questions of neuroscience posed above. We first study the role of sparse neural activity, as seen in the coding of sequential commands in a premotor area responsible for birdsong. We show that the sparse coding of temporal sequences in the songbird brain can, in a network where the feedforward plastic weights must translate the sparse sequential code into a time-varying muscle code, facilitate learning by minimizing synaptic interference. Next, we propose a biologically plausible synaptic plasticity rule that can perform goal-directed learning in recurrent networks of voltage-based spiking neurons that interact through conductances. Learning is based on the correlation of noisy local activity with a global reward signal; we prove that this rule performs stochastic gradient ascent on the reward. Thus, if the reward signal quantifies network performance on some desired task, the plasticity rule provably drives goal-directed learning in the network. To assess the convergence properties of the learning rule, we compare it with a known example of learning in the brain. Song-learning in finches is a clear example of a learned behavior, with detailed available neurophysiological data. With our learning rule, we train an anatomically accurate model birdsong network that drives a sound source to mimic an actual zebrafinch song. Simulation and

Hypothetical Pattern Recognition Design Using Multi-Layer Perceptorn Neural Network For Supervised Learning

Directory of Open Access Journals (Sweden)

Md. Abdullah-al-mamun

2015-08-01

Full Text Available Abstract Humans are capable to identifying diverse shape in the different pattern in the real world as effortless fashion due to their intelligence is grow since born with facing several learning process. Same way we can prepared an machine using human like brain called Artificial Neural Network that can be recognize different pattern from the real world object. Although the various techniques is exists to implementation the pattern recognition but recently the artificial neural network approaches have been giving the significant attention. Because the approached of artificial neural network is like a human brain that is learn from different observation and give a decision the previously learning rule. Over the 50 years research now a days pattern recognition for machine learning using artificial neural network got a significant achievement. For this reason many real world problem can be solve by modeling the pattern recognition process. The objective of this paper is to present the theoretical concept for pattern recognition design using Multi-Layer Perceptorn neural networkin the algorithm of artificial Intelligence as the best possible way of utilizing available resources to make a decision that can be a human like performance.

Deep learning classification in asteroseismology using an improved neural network

DEFF Research Database (Denmark)

Hon, Marc; Stello, Dennis; Yu, Jie

2018-01-01

Deep learning in the form of 1D convolutional neural networks have previously been shown to be capable of efficiently classifying the evolutionary state of oscillating red giants into red giant branch stars and helium-core burning stars by recognizing visual features in their asteroseismic...... frequency spectra. We elaborate further on the deep learning method by developing an improved convolutional neural network classifier. To make our method useful for current and future space missions such as K2, TESS, and PLATO, we train classifiers that are able to classify the evolutionary states of lower...

Biologically-inspired Learning in Pulsed Neural Networks

DEFF Research Database (Denmark)

Lehmann, Torsten; Woodburn, Robin

1999-01-01

Self-learning chips to implement many popular ANN (artificial neural network) algorithms are very difficult to design. We explain why this is so and say what lessons previous work teaches us in the design of self-learning systems. We offer a contribution to the `biologically-inspired' approach......, explaining what we mean by this term and providing an example of a robust, self-learning design that can solve simple classical-conditioning tasks. We give details of the design of individual circuits to perform component functions, which can then be combined into a network to solve the task. We argue...

Sensorimotor learning biases choice behavior: a learning neural field model for decision making.

Directory of Open Access Journals (Sweden)

Christian Klaes

Full Text Available According to a prominent view of sensorimotor processing in primates, selection and specification of possible actions are not sequential operations. Rather, a decision for an action emerges from competition between different movement plans, which are specified and selected in parallel. For action choices which are based on ambiguous sensory input, the frontoparietal sensorimotor areas are considered part of the common underlying neural substrate for selection and specification of action. These areas have been shown capable of encoding alternative spatial motor goals in parallel during movement planning, and show signatures of competitive value-based selection among these goals. Since the same network is also involved in learning sensorimotor associations, competitive action selection (decision making should not only be driven by the sensory evidence and expected reward in favor of either action, but also by the subject's learning history of different sensorimotor associations. Previous computational models of competitive neural decision making used predefined associations between sensory input and corresponding motor output. Such hard-wiring does not allow modeling of how decisions are influenced by sensorimotor learning or by changing reward contingencies. We present a dynamic neural field model which learns arbitrary sensorimotor associations with a reward-driven Hebbian learning algorithm. We show that the model accurately simulates the dynamics of action selection with different reward contingencies, as observed in monkey cortical recordings, and that it correctly predicted the pattern of choice errors in a control experiment. With our adaptive model we demonstrate how network plasticity, which is required for association learning and adaptation to new reward contingencies, can influence choice behavior. The field model provides an integrated and dynamic account for the operations of sensorimotor integration, working memory and action

NetTurnP – Neural Network Prediction of Beta-turns by Use of Evolutionary Information and Predicted Protein Sequence Features

DEFF Research Database (Denmark)

Petersen, Bent; Lundegaard, Claus; Petersen, Thomas Nordahl

2010-01-01

is the highest reported performance on a two-class prediction of β-turn and not-β-turn. Furthermore NetTurnP shows improved performance on some of the specific β-turn types. In the present work, neural network methods have been trained to predict β-turn or not and individual β-turn types from the primary amino......β-turns are the most common type of non-repetitive structures, and constitute on average 25% of the amino acids in proteins. The formation of β-turns plays an important role in protein folding, protein stability and molecular recognition processes. In this work we present the neural network method...... NetTurnP, for prediction of two-class β-turns and prediction of the individual β-turn types, by use of evolutionary information and predicted protein sequence features. It has been evaluated against a commonly used dataset BT426, and achieves a Matthews correlation coefficient of 0.50, which...

Ontology Mapping Neural Network: An Approach to Learning and Inferring Correspondences among Ontologies

Science.gov (United States)

Peng, Yefei

2010-01-01

An ontology mapping neural network (OMNN) is proposed in order to learn and infer correspondences among ontologies. It extends the Identical Elements Neural Network (IENN)'s ability to represent and map complex relationships. The learning dynamics of simultaneous (interlaced) training of similar tasks interact at the shared connections of the…

Alexnet Feature Extraction and Multi-Kernel Learning for Objectoriented Classification

Science.gov (United States)

Ding, L.; Li, H.; Hu, C.; Zhang, W.; Wang, S.

2018-04-01

In view of the fact that the deep convolutional neural network has stronger ability of feature learning and feature expression, an exploratory research is done on feature extraction and classification for high resolution remote sensing images. Taking the Google image with 0.3 meter spatial resolution in Ludian area of Yunnan Province as an example, the image segmentation object was taken as the basic unit, and the pre-trained AlexNet deep convolution neural network model was used for feature extraction. And the spectral features, AlexNet features and GLCM texture features are combined with multi-kernel learning and SVM classifier, finally the classification results were compared and analyzed. The results show that the deep convolution neural network can extract more accurate remote sensing image features, and significantly improve the overall accuracy of classification, and provide a reference value for earthquake disaster investigation and remote sensing disaster evaluation.

ALEXNET FEATURE EXTRACTION AND MULTI-KERNEL LEARNING FOR OBJECTORIENTED CLASSIFICATION

Directory of Open Access Journals (Sweden)

L. Ding

2018-04-01

Full Text Available In view of the fact that the deep convolutional neural network has stronger ability of feature learning and feature expression, an exploratory research is done on feature extraction and classification for high resolution remote sensing images. Taking the Google image with 0.3 meter spatial resolution in Ludian area of Yunnan Province as an example, the image segmentation object was taken as the basic unit, and the pre-trained AlexNet deep convolution neural network model was used for feature extraction. And the spectral features, AlexNet features and GLCM texture features are combined with multi-kernel learning and SVM classifier, finally the classification results were compared and analyzed. The results show that the deep convolution neural network can extract more accurate remote sensing image features, and significantly improve the overall accuracy of classification, and provide a reference value for earthquake disaster investigation and remote sensing disaster evaluation.

Manufacturing Scheduling Using Colored Petri Nets and Reinforcement Learning

Directory of Open Access Journals (Sweden)

Maria Drakaki

2017-02-01

Full Text Available Agent-based intelligent manufacturing control systems are capable to efficiently respond and adapt to environmental changes. Manufacturing system adaptation and evolution can be addressed with learning mechanisms that increase the intelligence of agents. In this paper a manufacturing scheduling method is presented based on Timed Colored Petri Nets (CTPNs and reinforcement learning (RL. CTPNs model the manufacturing system and implement the scheduling. In the search for an optimal solution a scheduling agent uses RL and in particular the Q-learning algorithm. A warehouse order-picking scheduling is presented as a case study to illustrate the method. The proposed scheduling method is compared to existing methods. Simulation and state space results are used to evaluate performance and identify system properties.

Learning by stimulation avoidance: A principle to control spiking neural networks dynamics.

Science.gov (United States)

Sinapayen, Lana; Masumori, Atsushi; Ikegami, Takashi

2017-01-01

Learning based on networks of real neurons, and learning based on biologically inspired models of neural networks, have yet to find general learning rules leading to widespread applications. In this paper, we argue for the existence of a principle allowing to steer the dynamics of a biologically inspired neural network. Using carefully timed external stimulation, the network can be driven towards a desired dynamical state. We term this principle "Learning by Stimulation Avoidance" (LSA). We demonstrate through simulation that the minimal sufficient conditions leading to LSA in artificial networks are also sufficient to reproduce learning results similar to those obtained in biological neurons by Shahaf and Marom, and in addition explains synaptic pruning. We examined the underlying mechanism by simulating a small network of 3 neurons, then scaled it up to a hundred neurons. We show that LSA has a higher explanatory power than existing hypotheses about the response of biological neural networks to external simulation, and can be used as a learning rule for an embodied application: learning of wall avoidance by a simulated robot. In other works, reinforcement learning with spiking networks can be obtained through global reward signals akin simulating the dopamine system; we believe that this is the first project demonstrating sensory-motor learning with random spiking networks through Hebbian learning relying on environmental conditions without a separate reward system.

Lifelong learning of human actions with deep neural network self-organization.

Science.gov (United States)

Parisi, German I; Tani, Jun; Weber, Cornelius; Wermter, Stefan

2017-12-01

Lifelong learning is fundamental in autonomous robotics for the acquisition and fine-tuning of knowledge through experience. However, conventional deep neural models for action recognition from videos do not account for lifelong learning but rather learn a batch of training data with a predefined number of action classes and samples. Thus, there is the need to develop learning systems with the ability to incrementally process available perceptual cues and to adapt their responses over time. We propose a self-organizing neural architecture for incrementally learning to classify human actions from video sequences. The architecture comprises growing self-organizing networks equipped with recurrent neurons for processing time-varying patterns. We use a set of hierarchically arranged recurrent networks for the unsupervised learning of action representations with increasingly large spatiotemporal receptive fields. Lifelong learning is achieved in terms of prediction-driven neural dynamics in which the growth and the adaptation of the recurrent networks are driven by their capability to reconstruct temporally ordered input sequences. Experimental results on a classification task using two action benchmark datasets show that our model is competitive with state-of-the-art methods for batch learning also when a significant number of sample labels are missing or corrupted during training sessions. Additional experiments show the ability of our model to adapt to non-stationary input avoiding catastrophic interference. Copyright © 2017 The Author(s). Published by Elsevier Ltd.. All rights reserved.

A theory of causal learning in children: Causal maps and Bayes nets

OpenAIRE

Gopnik, A; Glymour, C; Sobel, D M; Schulz, L E; Kushnir, T; Danks, D

2004-01-01

The authors outline a cognitive and computational account of causal learning in children. They propose that children use specialized cognitive systems that allow them to recover an accurate "causal map" of the world: an abstract, coherent, learned representation of the causal relations among events. This kind of knowledge can be perspicuously understood in terms of the formalism of directed graphical causal models, or Bayes nets. Children's causal learning and inference may involve computatio...

Neural Control of a Tracking Task via Attention-Gated Reinforcement Learning for Brain-Machine Interfaces.

Science.gov (United States)

Wang, Yiwen; Wang, Fang; Xu, Kai; Zhang, Qiaosheng; Zhang, Shaomin; Zheng, Xiaoxiang

2015-05-01

Reinforcement learning (RL)-based brain machine interfaces (BMIs) enable the user to learn from the environment through interactions to complete the task without desired signals, which is promising for clinical applications. Previous studies exploited Q-learning techniques to discriminate neural states into simple directional actions providing the trial initial timing. However, the movements in BMI applications can be quite complicated, and the action timing explicitly shows the intention when to move. The rich actions and the corresponding neural states form a large state-action space, imposing generalization difficulty on Q-learning. In this paper, we propose to adopt attention-gated reinforcement learning (AGREL) as a new learning scheme for BMIs to adaptively decode high-dimensional neural activities into seven distinct movements (directional moves, holdings and resting) due to the efficient weight-updating. We apply AGREL on neural data recorded from M1 of a monkey to directly predict a seven-action set in a time sequence to reconstruct the trajectory of a center-out task. Compared to Q-learning techniques, AGREL could improve the target acquisition rate to 90.16% in average with faster convergence and more stability to follow neural activity over multiple days, indicating the potential to achieve better online decoding performance for more complicated BMI tasks.

Bio-inspired Artificial Intelligence: А Generalized Net Model of the Regularization Process in MLP

Directory of Open Access Journals (Sweden)

Stanimir Surchev

2013-10-01

Full Text Available Many objects and processes inspired by the nature have been recreated by the scientists. The inspiration to create a Multilayer Neural Network came from human brain as member of the group. It possesses complicated structure and it is difficult to recreate, because of the existence of too many processes that require different solving methods. The aim of the following paper is to describe one of the methods that improve learning process of Artificial Neural Network. The proposed generalized net method presents Regularization process in Multilayer Neural Network. The purpose of verification is to protect the neural network from overfitting. The regularization is commonly used in neural network training process. Many methods of verification are present, the subject of interest is the one known as Regularization. It contains function in order to set weights and biases with smaller values to protect from overfitting.

Learning Perfectly Secure Cryptography to Protect Communications with Adversarial Neural Cryptography

Directory of Open Access Journals (Sweden)

Murilo Coutinho

2018-04-01

Full Text Available Researches in Artificial Intelligence (AI have achieved many important breakthroughs, especially in recent years. In some cases, AI learns alone from scratch and performs human tasks faster and better than humans. With the recent advances in AI, it is natural to wonder whether Artificial Neural Networks will be used to successfully create or break cryptographic algorithms. Bibliographic review shows the main approach to this problem have been addressed throughout complex Neural Networks, but without understanding or proving the security of the generated model. This paper presents an analysis of the security of cryptographic algorithms generated by a new technique called Adversarial Neural Cryptography (ANC. Using the proposed network, we show limitations and directions to improve the current approach of ANC. Training the proposed Artificial Neural Network with the improved model of ANC, we show that artificially intelligent agents can learn the unbreakable One-Time Pad (OTP algorithm, without human knowledge, to communicate securely through an insecure communication channel. This paper shows in which conditions an AI agent can learn a secure encryption scheme. However, it also shows that, without a stronger adversary, it is more likely to obtain an insecure one.

Learning Perfectly Secure Cryptography to Protect Communications with Adversarial Neural Cryptography.

Science.gov (United States)

Coutinho, Murilo; de Oliveira Albuquerque, Robson; Borges, Fábio; García Villalba, Luis Javier; Kim, Tai-Hoon

2018-04-24

Researches in Artificial Intelligence (AI) have achieved many important breakthroughs, especially in recent years. In some cases, AI learns alone from scratch and performs human tasks faster and better than humans. With the recent advances in AI, it is natural to wonder whether Artificial Neural Networks will be used to successfully create or break cryptographic algorithms. Bibliographic review shows the main approach to this problem have been addressed throughout complex Neural Networks, but without understanding or proving the security of the generated model. This paper presents an analysis of the security of cryptographic algorithms generated by a new technique called Adversarial Neural Cryptography (ANC). Using the proposed network, we show limitations and directions to improve the current approach of ANC. Training the proposed Artificial Neural Network with the improved model of ANC, we show that artificially intelligent agents can learn the unbreakable One-Time Pad (OTP) algorithm, without human knowledge, to communicate securely through an insecure communication channel. This paper shows in which conditions an AI agent can learn a secure encryption scheme. However, it also shows that, without a stronger adversary, it is more likely to obtain an insecure one.

Introduction to spiking neural networks: Information processing, learning and applications.

Science.gov (United States)

Ponulak, Filip; Kasinski, Andrzej

2011-01-01

The concept that neural information is encoded in the firing rate of neurons has been the dominant paradigm in neurobiology for many years. This paradigm has also been adopted by the theory of artificial neural networks. Recent physiological experiments demonstrate, however, that in many parts of the nervous system, neural code is founded on the timing of individual action potentials. This finding has given rise to the emergence of a new class of neural models, called spiking neural networks. In this paper we summarize basic properties of spiking neurons and spiking networks. Our focus is, specifically, on models of spike-based information coding, synaptic plasticity and learning. We also survey real-life applications of spiking models. The paper is meant to be an introduction to spiking neural networks for scientists from various disciplines interested in spike-based neural processing.

A learning algorithm for oscillatory cellular neural networks.

Science.gov (United States)

Ho, C Y.; Kurokawa, H

1999-07-01

We present a cellular type oscillatory neural network for temporal segregation of stationary input patterns. The model comprises an array of locally connected neural oscillators with connections limited to a 4-connected neighborhood. The architecture is reminiscent of the well-known cellular neural network that consists of local connection for feature extraction. By means of a novel learning rule and an initialization scheme, global synchronization can be accomplished without incurring any erroneous synchrony among uncorrelated objects. Each oscillator comprises two mutually coupled neurons, and neurons share a piecewise-linear activation function characteristic. The dynamics of traditional oscillatory models is simplified by using only one plastic synapse, and the overall complexity for hardware implementation is reduced. Based on the connectedness of image segments, it is shown that global synchronization and desynchronization can be achieved by means of locally connected synapses, and this opens up a tremendous application potential for the proposed architecture. Furthermore, by using special grouping synapses it is demonstrated that temporal segregation of overlapping gray-level and color segments can also be achieved. Finally, simulation results show that the learning rule proposed circumvents the problem of component mismatches, and hence facilitates a large-scale integration.

Learning and forgetting on asymmetric, diluted neural networks

International Nuclear Information System (INIS)

Derrida, B.; Nadal, J.P.

1987-01-01

It is possible to construct diluted asymmetric models of neural networks for which the dynamics can be calculated exactly. The authors test several learning schemes, in particular, models for which the values of the synapses remain bounded and depend on the history. Our analytical results on the relative efficiencies of the various learning schemes are qualitatively similar to the corresponding ones obtained numerically on fully connected symmetric networks

Strategies influence neural activity for feedback learning across child and adolescent development.

Science.gov (United States)

Peters, Sabine; Koolschijn, P Cédric M P; Crone, Eveline A; Van Duijvenvoorde, Anna C K; Raijmakers, Maartje E J

2014-09-01

Learning from feedback is an important aspect of executive functioning that shows profound improvements during childhood and adolescence. This is accompanied by neural changes in the feedback-learning network, which includes pre-supplementary motor area (pre- SMA)/anterior cingulate cortex (ACC), dorsolateral prefrontal cortex (DLPFC), superior parietal cortex (SPC), and the basal ganglia. However, there can be considerable differences within age ranges in performance that are ascribed to differences in strategy use. This is problematic for traditional approaches of analyzing developmental data, in which age groups are assumed to be homogenous in strategy use. In this study, we used latent variable models to investigate if underlying strategy groups could be detected for a feedback-learning task and whether there were differences in neural activation patterns between strategies. In a sample of 268 participants between ages 8 to 25 years, we observed four underlying strategy groups, which were cut across age groups and varied in the optimality of executive functioning. These strategy groups also differed in neural activity during learning; especially the most optimal performing group showed more activity in DLPFC, SPC and pre-SMA/ACC compared to the other groups. However, age differences remained an important contributor to neural activation, even when correcting for strategy. These findings contribute to the debate of age versus performance predictors of neural development, and highlight the importance of studying individual differences in strategy use when studying development. Copyright © 2014 Elsevier Ltd. All rights reserved.

Application and Simulation of Fuzzy Neural Network PID Controller in the Aircraft Cabin Temperature

Directory of Open Access Journals (Sweden)

Ding Fang

2013-06-01

Full Text Available Considering complex factors of affecting ambient temperature in Aircraft cabin, and some shortages of traditional PID control like the parameters difficult to be tuned and control ineffective, this paper puts forward the intelligent PID algorithm that makes fuzzy logic method and neural network together, scheming out the fuzzy neural net PID controller. After the correction of the fuzzy inference and dynamic learning of neural network, PID parameters of the controller get the optimal parameters. MATLAB simulation results of the cabin temperature control model show that the performance of the fuzzy neural network PID controller has been greatly improved, with faster response, smaller overshoot and better adaptability.

Improved Discriminability of Spatiotemporal Neural Patterns in Rat Motor Cortical Areas as Directional Choice Learning Progresses

Directory of Open Access Journals (Sweden)

Hongwei eMao

2015-03-01

Full Text Available Animals learn to choose a proper action among alternatives to improve their odds of success in food foraging and other activities critical for survival. Through trial-and-error, they learn correct associations between their choices and external stimuli. While a neural network that underlies such learning process has been identified at a high level, it is still unclear how individual neurons and a neural ensemble adapt as learning progresses. In this study, we monitored the activity of single units in the rat medial and lateral agranular (AGm and AGl, respectively areas as rats learned to make a left or right side lever press in response to a left or right side light cue. We noticed that rat movement parameters during the performance of the directional choice task quickly became stereotyped during the first 2-3 days or sessions. But learning the directional choice problem took weeks to occur. Accompanying rats’ behavioral performance adaptation, we observed neural modulation by directional choice in recorded single units. Our analysis shows that ensemble mean firing rates in the cue-on period did not change significantly as learning progressed, and the ensemble mean rate difference between left and right side choices did not show a clear trend of change either. However, the spatiotemporal firing patterns of the neural ensemble exhibited improved discriminability between the two directional choices through learning. These results suggest a spatiotemporal neural coding scheme in a motor cortical neural ensemble that may be responsible for and contributing to learning the directional choice task.

Adaptive competitive learning neural networks

Directory of Open Access Journals (Sweden)

Ahmed R. Abas

2013-11-01

Full Text Available In this paper, the adaptive competitive learning (ACL neural network algorithm is proposed. This neural network not only groups similar input feature vectors together but also determines the appropriate number of groups of these vectors. This algorithm uses a new proposed criterion referred to as the ACL criterion. This criterion evaluates different clustering structures produced by the ACL neural network for an input data set. Then, it selects the best clustering structure and the corresponding network architecture for this data set. The selected structure is composed of the minimum number of clusters that are compact and balanced in their sizes. The selected network architecture is efficient, in terms of its complexity, as it contains the minimum number of neurons. Synaptic weight vectors of these neurons represent well-separated, compact and balanced clusters in the input data set. The performance of the ACL algorithm is evaluated and compared with the performance of a recently proposed algorithm in the literature in clustering an input data set and determining its number of clusters. Results show that the ACL algorithm is more accurate and robust in both determining the number of clusters and allocating input feature vectors into these clusters than the other algorithm especially with data sets that are sparsely distributed.

Using CSCW for developing problem-oriented teaching and learning in a net environment

DEFF Research Database (Denmark)

Cheesman, Robin; Heilesen, Simon

Roskilde University’s master’s programme in computer-mediated communication combines face-to-face seminars with net seminars focusing on collaborative project work. Net-based learning based on CSCW offers both advantages and pitfalls: (i) it helps to activate all students, (ii) it fosters...... complexity in organising tasks, (iii) asynchronous environment generates a need for synchronous communication, and (iv) exaggerated structuring limits self-organising and motivation....

NetBeans IDE 8 cookbook

CERN Document Server

Salter, David

2014-01-01

If you're a Java developer of any level using NetBeans and want to learn how to get the most out of NetBeans, then this book is for you. Learning how to utilize NetBeans will provide a firm foundation for your Java application development.

Computation and control with neural nets

Energy Technology Data Exchange (ETDEWEB)

Corneliusen, A.; Terdal, P.; Knight, T.; Spencer, J.

1989-10-04

As energies have increased exponentially with time so have the size and complexity of accelerators and control systems. NN may offer the kinds of improvements in computation and control that are needed to maintain acceptable functionality. For control their associative characteristics could provide signal conversion or data translation. Because they can do any computation such as least squares, they can close feedback loops autonomously to provide intelligent control at the point of action rather than at a central location that requires transfers, conversions, hand-shaking and other costly repetitions like input protection. Both computation and control can be integrated on a single chip, printed circuit or an optical equivalent that is also inherently faster through full parallel operation. For such reasons one expects lower costs and better results. Such systems could be optimized by integrating sensor and signal processing functions. Distributed nets of such hardware could communicate and provide global monitoring and multiprocessing in various ways e.g. via token, slotted or parallel rings (or Steiner trees) for compatibility with existing systems. Problems and advantages of this approach such as an optimal, real-time Turing machine are discussed. Simple examples are simulated and hardware implemented using discrete elements that demonstrate some basic characteristics of learning and parallelism. Future microprocessors' are predicted and requested on this basis. 19 refs., 18 figs.

Computation and control with neural nets

International Nuclear Information System (INIS)

Corneliusen, A.; Terdal, P.; Knight, T.; Spencer, J.

1989-01-01

As energies have increased exponentially with time so have the size and complexity of accelerators and control systems. NN may offer the kinds of improvements in computation and control that are needed to maintain acceptable functionality. For control their associative characteristics could provide signal conversion or data translation. Because they can do any computation such as least squares, they can close feedback loops autonomously to provide intelligent control at the point of action rather than at a central location that requires transfers, conversions, hand-shaking and other costly repetitions like input protection. Both computation and control can be integrated on a single chip, printed circuit or an optical equivalent that is also inherently faster through full parallel operation. For such reasons one expects lower costs and better results. Such systems could be optimized by integrating sensor and signal processing functions. Distributed nets of such hardware could communicate and provide global monitoring and multiprocessing in various ways e.g. via token, slotted or parallel rings (or Steiner trees) for compatibility with existing systems. Problems and advantages of this approach such as an optimal, real-time Turing machine are discussed. Simple examples are simulated and hardware implemented using discrete elements that demonstrate some basic characteristics of learning and parallelism. Future 'microprocessors' are predicted and requested on this basis. 19 refs., 18 figs

Nuclear power plant monitoring using real-time learning neural network

International Nuclear Information System (INIS)

Nabeshima, Kunihiko; Tuerkcan, E.; Ciftcioglu, O.

1994-01-01

In the present research, artificial neural network (ANN) with real-time adaptive learning is developed for the plant wide monitoring of Borssele Nuclear Power Plant (NPP). Adaptive ANN learning capability is integrated to the monitoring system so that robust and sensitive on-line monitoring is achieved in real-time environment. The major advantages provided by ANN are that system modelling is formed by means of measurement information obtained from a multi-output process system, explicit modelling is not required and the modelling is not restricted to linear systems. Also ANN can respond very fast to anomalous operational conditions. The real-time ANN learning methodology with adaptive real-time monitoring capability is described below for the wide-range and plant-wide data from an operating nuclear power plant. The layered neural network with error backpropagation algorithm for learning has three layers. The network type is auto-associative, inputs and outputs are exactly the same, using 12 plant signals. (author)

Robust Control Methods for On-Line Statistical Learning

Directory of Open Access Journals (Sweden)

Capobianco Enrico

2001-01-01

Full Text Available The issue of controlling that data processing in an experiment results not affected by the presence of outliers is relevant for statistical control and learning studies. Learning schemes should thus be tested for their capacity of handling outliers in the observed training set so to achieve reliable estimates with respect to the crucial bias and variance aspects. We describe possible ways of endowing neural networks with statistically robust properties by defining feasible error criteria. It is convenient to cast neural nets in state space representations and apply both Kalman filter and stochastic approximation procedures in order to suggest statistically robustified solutions for on-line learning.

A Self-Organizing Incremental Neural Network based on local distribution learning.

Science.gov (United States)

Xing, Youlu; Shi, Xiaofeng; Shen, Furao; Zhou, Ke; Zhao, Jinxi

2016-12-01

In this paper, we propose an unsupervised incremental learning neural network based on local distribution learning, which is called Local Distribution Self-Organizing Incremental Neural Network (LD-SOINN). The LD-SOINN combines the advantages of incremental learning and matrix learning. It can automatically discover suitable nodes to fit the learning data in an incremental way without a priori knowledge such as the structure of the network. The nodes of the network store rich local information regarding the learning data. The adaptive vigilance parameter guarantees that LD-SOINN is able to add new nodes for new knowledge automatically and the number of nodes will not grow unlimitedly. While the learning process continues, nodes that are close to each other and have similar principal components are merged to obtain a concise local representation, which we call a relaxation data representation. A denoising process based on density is designed to reduce the influence of noise. Experiments show that the LD-SOINN performs well on both artificial and real-word data. Copyright © 2016 Elsevier Ltd. All rights reserved.

NetTurnP--neural network prediction of beta-turns by use of evolutionary information and predicted protein sequence features.

Directory of Open Access Journals (Sweden)

Bent Petersen

Full Text Available UNLABELLED: β-turns are the most common type of non-repetitive structures, and constitute on average 25% of the amino acids in proteins. The formation of β-turns plays an important role in protein folding, protein stability and molecular recognition processes. In this work we present the neural network method NetTurnP, for prediction of two-class β-turns and prediction of the individual β-turn types, by use of evolutionary information and predicted protein sequence features. It has been evaluated against a commonly used dataset BT426, and achieves a Matthews correlation coefficient of 0.50, which is the highest reported performance on a two-class prediction of β-turn and not-β-turn. Furthermore NetTurnP shows improved performance on some of the specific β-turn types. In the present work, neural network methods have been trained to predict β-turn or not and individual β-turn types from the primary amino acid sequence. The individual β-turn types I, I', II, II', VIII, VIa1, VIa2, VIba and IV have been predicted based on classifications by PROMOTIF, and the two-class prediction of β-turn or not is a superset comprised of all β-turn types. The performance is evaluated using a golden set of non-homologous sequences known as BT426. Our two-class prediction method achieves a performance of: MCC=0.50, Qtotal=82.1%, sensitivity=75.6%, PPV=68.8% and AUC=0.864. We have compared our performance to eleven other prediction methods that obtain Matthews correlation coefficients in the range of 0.17-0.47. For the type specific β-turn predictions, only type I and II can be predicted with reasonable Matthews correlation coefficients, where we obtain performance values of 0.36 and 0.31, respectively. CONCLUSION: The NetTurnP method has been implemented as a webserver, which is freely available at http://www.cbs.dtu.dk/services/NetTurnP/. NetTurnP is the only available webserver that allows submission of multiple sequences.

NetTurnP--neural network prediction of beta-turns by use of evolutionary information and predicted protein sequence features.

Science.gov (United States)

Petersen, Bent; Lundegaard, Claus; Petersen, Thomas Nordahl

2010-11-30

β-turns are the most common type of non-repetitive structures, and constitute on average 25% of the amino acids in proteins. The formation of β-turns plays an important role in protein folding, protein stability and molecular recognition processes. In this work we present the neural network method NetTurnP, for prediction of two-class β-turns and prediction of the individual β-turn types, by use of evolutionary information and predicted protein sequence features. It has been evaluated against a commonly used dataset BT426, and achieves a Matthews correlation coefficient of 0.50, which is the highest reported performance on a two-class prediction of β-turn and not-β-turn. Furthermore NetTurnP shows improved performance on some of the specific β-turn types. In the present work, neural network methods have been trained to predict β-turn or not and individual β-turn types from the primary amino acid sequence. The individual β-turn types I, I', II, II', VIII, VIa1, VIa2, VIba and IV have been predicted based on classifications by PROMOTIF, and the two-class prediction of β-turn or not is a superset comprised of all β-turn types. The performance is evaluated using a golden set of non-homologous sequences known as BT426. Our two-class prediction method achieves a performance of: MCC=0.50, Qtotal=82.1%, sensitivity=75.6%, PPV=68.8% and AUC=0.864. We have compared our performance to eleven other prediction methods that obtain Matthews correlation coefficients in the range of 0.17-0.47. For the type specific β-turn predictions, only type I and II can be predicted with reasonable Matthews correlation coefficients, where we obtain performance values of 0.36 and 0.31, respectively. The NetTurnP method has been implemented as a webserver, which is freely available at http://www.cbs.dtu.dk/services/NetTurnP/. NetTurnP is the only available webserver that allows submission of multiple sequences.

A novel Bayesian learning method for information aggregation in modular neural networks

DEFF Research Database (Denmark)

Wang, Pan; Xu, Lida; Zhou, Shang-Ming

2010-01-01

Modular neural network is a popular neural network model which has many successful applications. In this paper, a sequential Bayesian learning (SBL) is proposed for modular neural networks aiming at efficiently aggregating the outputs of members of the ensemble. The experimental results on eight...... benchmark problems have demonstrated that the proposed method can perform information aggregation efficiently in data modeling....

Neural prediction errors reveal a risk-sensitive reinforcement-learning process in the human brain.

Science.gov (United States)

Niv, Yael; Edlund, Jeffrey A; Dayan, Peter; O'Doherty, John P

2012-01-11

Humans and animals are exquisitely, though idiosyncratically, sensitive to risk or variance in the outcomes of their actions. Economic, psychological, and neural aspects of this are well studied when information about risk is provided explicitly. However, we must normally learn about outcomes from experience, through trial and error. Traditional models of such reinforcement learning focus on learning about the mean reward value of cues and ignore higher order moments such as variance. We used fMRI to test whether the neural correlates of human reinforcement learning are sensitive to experienced risk. Our analysis focused on anatomically delineated regions of a priori interest in the nucleus accumbens, where blood oxygenation level-dependent (BOLD) signals have been suggested as correlating with quantities derived from reinforcement learning. We first provide unbiased evidence that the raw BOLD signal in these regions corresponds closely to a reward prediction error. We then derive from this signal the learned values of cues that predict rewards of equal mean but different variance and show that these values are indeed modulated by experienced risk. Moreover, a close neurometric-psychometric coupling exists between the fluctuations of the experience-based evaluations of risky options that we measured neurally and the fluctuations in behavioral risk aversion. This suggests that risk sensitivity is integral to human learning, illuminating economic models of choice, neuroscientific models of affective learning, and the workings of the underlying neural mechanisms.

Parallelization of learning problems by artificial neural networks. Application in external radiotherapy

International Nuclear Information System (INIS)

Sauget, M.

2007-12-01

This research is about the application of neural networks used in the external radiotherapy domain. The goal is to elaborate a new evaluating system for the radiation dose distributions in heterogeneous environments. The al objective of this work is to build a complete tool kit to evaluate the optimal treatment planning. My st research point is about the conception of an incremental learning algorithm. The interest of my work is to combine different optimizations specialized in the function interpolation and to propose a new algorithm allowing to change the neural network architecture during the learning phase. This algorithm allows to minimise the al size of the neural network while keeping a good accuracy. The second part of my research is to parallelize the previous incremental learning algorithm. The goal of that work is to increase the speed of the learning step as well as the size of the learned dataset needed in a clinical case. For that, our incremental learning algorithm presents an original data decomposition with overlapping, together with a fault tolerance mechanism. My last research point is about a fast and accurate algorithm computing the radiation dose deposit in any heterogeneous environment. At the present time, the existing solutions used are not optimal. The fast solution are not accurate and do not give an optimal treatment planning. On the other hand, the accurate solutions are far too slow to be used in a clinical context. Our algorithm answers to this problem by bringing rapidity and accuracy. The concept is to use a neural network adequately learned together with a mechanism taking into account the environment changes. The advantages of this algorithm is to avoid the use of a complex physical code while keeping a good accuracy and reasonable computation times. (author)

High-Performance Neural Networks for Visual Object Classification

OpenAIRE

Cireşan, Dan C.; Meier, Ueli; Masci, Jonathan; Gambardella, Luca M.; Schmidhuber, Jürgen

2011-01-01

We present a fast, fully parameterizable GPU implementation of Convolutional Neural Network variants. Our feature extractors are neither carefully designed nor pre-wired, but rather learned in a supervised way. Our deep hierarchical architectures achieve the best published results on benchmarks for object classification (NORB, CIFAR10) and handwritten digit recognition (MNIST), with error rates of 2.53%, 19.51%, 0.35%, respectively. Deep nets trained by simple back-propagation perform better ...

Neural correlates of context-dependent feature conjunction learning in visual search tasks.

Science.gov (United States)

Reavis, Eric A; Frank, Sebastian M; Greenlee, Mark W; Tse, Peter U

2016-06-01

Many perceptual learning experiments show that repeated exposure to a basic visual feature such as a specific orientation or spatial frequency can modify perception of that feature, and that those perceptual changes are associated with changes in neural tuning early in visual processing. Such perceptual learning effects thus exert a bottom-up influence on subsequent stimulus processing, independent of task-demands or endogenous influences (e.g., volitional attention). However, it is unclear whether such bottom-up changes in perception can occur as more complex stimuli such as conjunctions of visual features are learned. It is not known whether changes in the efficiency with which people learn to process feature conjunctions in a task (e.g., visual search) reflect true bottom-up perceptual learning versus top-down, task-related learning (e.g., learning better control of endogenous attention). Here we show that feature conjunction learning in visual search leads to bottom-up changes in stimulus processing. First, using fMRI, we demonstrate that conjunction learning in visual search has a distinct neural signature: an increase in target-evoked activity relative to distractor-evoked activity (i.e., a relative increase in target salience). Second, we demonstrate that after learning, this neural signature is still evident even when participants passively view learned stimuli while performing an unrelated, attention-demanding task. This suggests that conjunction learning results in altered bottom-up perceptual processing of the learned conjunction stimuli (i.e., a perceptual change independent of the task). We further show that the acquired change in target-evoked activity is contextually dependent on the presence of distractors, suggesting that search array Gestalts are learned. Hum Brain Mapp 37:2319-2330, 2016. © 2016 Wiley Periodicals, Inc. © 2016 Wiley Periodicals, Inc.

Neural Correlates of High Performance in Foreign Language Vocabulary Learning

Science.gov (United States)

Macedonia, Manuela; Muller, Karsten; Friederici, Angela D.

2010-01-01

Learning vocabulary in a foreign language is a laborious task which people perform with varying levels of success. Here, we investigated the neural underpinning of high performance on this task. In a within-subjects paradigm, participants learned 92 vocabulary items under two multimodal conditions: one condition paired novel words with iconic…

Neural network signal understanding for instrumentation

DEFF Research Database (Denmark)

Pau, L. F.; Johansen, F. S.

1990-01-01

understanding research is surveyed, and the selected implementation and its performance in terms of correct classification rates and robustness to noise are described. Formal results on neural net training time and sensitivity to weights are given. A theory for neural control using functional link nets is given...

Learning characteristics of a space-time neural network as a tether skiprope observer

Science.gov (United States)

Lea, Robert N.; Villarreal, James A.; Jani, Yashvant; Copeland, Charles

1993-01-01

The Software Technology Laboratory at the Johnson Space Center is testing a Space Time Neural Network (STNN) for observing tether oscillations present during retrieval of a tethered satellite. Proper identification of tether oscillations, known as 'skiprope' motion, is vital to safe retrieval of the tethered satellite. Our studies indicate that STNN has certain learning characteristics that must be understood properly to utilize this type of neural network for the tethered satellite problem. We present our findings on the learning characteristics including a learning rate versus momentum performance table.

Deep neural networks for direct, featureless learning through observation: The case of two-dimensional spin models

Science.gov (United States)

Mills, Kyle; Tamblyn, Isaac

2018-03-01

We demonstrate the capability of a convolutional deep neural network in predicting the nearest-neighbor energy of the 4 ×4 Ising model. Using its success at this task, we motivate the study of the larger 8 ×8 Ising model, showing that the deep neural network can learn the nearest-neighbor Ising Hamiltonian after only seeing a vanishingly small fraction of configuration space. Additionally, we show that the neural network has learned both the energy and magnetization operators with sufficient accuracy to replicate the low-temperature Ising phase transition. We then demonstrate the ability of the neural network to learn other spin models, teaching the convolutional deep neural network to accurately predict the long-range interaction of a screened Coulomb Hamiltonian, a sinusoidally attenuated screened Coulomb Hamiltonian, and a modified Potts model Hamiltonian. In the case of the long-range interaction, we demonstrate the ability of the neural network to recover the phase transition with equivalent accuracy to the numerically exact method. Furthermore, in the case of the long-range interaction, the benefits of the neural network become apparent; it is able to make predictions with a high degree of accuracy, and do so 1600 times faster than a CUDA-optimized exact calculation. Additionally, we demonstrate how the neural network succeeds at these tasks by looking at the weights learned in a simplified demonstration.

QSAR modelling using combined simple competitive learning networks and RBF neural networks.

Science.gov (United States)

Sheikhpour, R; Sarram, M A; Rezaeian, M; Sheikhpour, E

2018-04-01

The aim of this study was to propose a QSAR modelling approach based on the combination of simple competitive learning (SCL) networks with radial basis function (RBF) neural networks for predicting the biological activity of chemical compounds. The proposed QSAR method consisted of two phases. In the first phase, an SCL network was applied to determine the centres of an RBF neural network. In the second phase, the RBF neural network was used to predict the biological activity of various phenols and Rho kinase (ROCK) inhibitors. The predictive ability of the proposed QSAR models was evaluated and compared with other QSAR models using external validation. The results of this study showed that the proposed QSAR modelling approach leads to better performances than other models in predicting the biological activity of chemical compounds. This indicated the efficiency of simple competitive learning networks in determining the centres of RBF neural networks.

Deep learning in TMVA Benchmarking Benchmarking TMVA DNN Integration of a Deep Autoencoder

CERN Document Server

Huwiler, Marc

2017-01-01

The TMVA library in ROOT is dedicated to multivariate analysis, and in partic- ular oers numerous machine learning algorithms in a standardized framework. It is widely used in High Energy Physics for data analysis, mainly to perform regression and classication. To keep up to date with the state of the art in deep learning, a new deep learning module was being developed this summer, oering deep neural net- work, convolutional neural network, and autoencoder. TMVA did not have yet any autoencoder method, and the present project consists in implementing the TMVA autoencoder class based on the deep learning module. It also includes some bench- marking performed on the actual deep neural network implementation, in comparison to the Keras framework with Tensorflow and Theano backend.

Saliency U-Net: A regional saliency map-driven hybrid deep learning network for anomaly segmentation

Science.gov (United States)

Karargyros, Alex; Syeda-Mahmood, Tanveer

2018-02-01

Deep learning networks are gaining popularity in many medical image analysis tasks due to their generalized ability to automatically extract relevant features from raw images. However, this can make the learning problem unnecessarily harder requiring network architectures of high complexity. In case of anomaly detection, in particular, there is often sufficient regional difference between the anomaly and the surrounding parenchyma that could be easily highlighted through bottom-up saliency operators. In this paper we propose a new hybrid deep learning network using a combination of raw image and such regional maps to more accurately learn the anomalies using simpler network architectures. Specifically, we modify a deep learning network called U-Net using both the raw and pre-segmented images as input to produce joint encoding (contraction) and expansion paths (decoding) in the U-Net. We present results of successfully delineating subdural and epidural hematomas in brain CT imaging and liver hemangioma in abdominal CT images using such network.

Real-time cerebellar neuroprosthetic system based on a spiking neural network model of motor learning.

Science.gov (United States)

Xu, Tao; Xiao, Na; Zhai, Xiaolong; Kwan Chan, Pak; Tin, Chung

2018-02-01

Damage to the brain, as a result of various medical conditions, impacts the everyday life of patients and there is still no complete cure to neurological disorders. Neuroprostheses that can functionally replace the damaged neural circuit have recently emerged as a possible solution to these problems. Here we describe the development of a real-time cerebellar neuroprosthetic system to substitute neural function in cerebellar circuitry for learning delay eyeblink conditioning (DEC). The system was empowered by a biologically realistic spiking neural network (SNN) model of the cerebellar neural circuit, which considers the neuronal population and anatomical connectivity of the network. The model simulated synaptic plasticity critical for learning DEC. This SNN model was carefully implemented on a field programmable gate array (FPGA) platform for real-time simulation. This hardware system was interfaced in in vivo experiments with anesthetized rats and it used neural spikes recorded online from the animal to learn and trigger conditioned eyeblink in the animal during training. This rat-FPGA hybrid system was able to process neuronal spikes in real-time with an embedded cerebellum model of ~10 000 neurons and reproduce learning of DEC with different inter-stimulus intervals. Our results validated that the system performance is physiologically relevant at both the neural (firing pattern) and behavioral (eyeblink pattern) levels. This integrated system provides the sufficient computation power for mimicking the cerebellar circuit in real-time. The system interacts with the biological system naturally at the spike level and can be generalized for including other neural components (neuron types and plasticity) and neural functions for potential neuroprosthetic applications.

Opponent appetitive-aversive neural processes underlie predictive learning of pain relief.

Science.gov (United States)

Seymour, Ben; O'Doherty, John P; Koltzenburg, Martin; Wiech, Katja; Frackowiak, Richard; Friston, Karl; Dolan, Raymond

2005-09-01

Termination of a painful or unpleasant event can be rewarding. However, whether the brain treats relief in a similar way as it treats natural reward is unclear, and the neural processes that underlie its representation as a motivational goal remain poorly understood. We used fMRI (functional magnetic resonance imaging) to investigate how humans learn to generate expectations of pain relief. Using a pavlovian conditioning procedure, we show that subjects experiencing prolonged experimentally induced pain can be conditioned to predict pain relief. This proceeds in a manner consistent with contemporary reward-learning theory (average reward/loss reinforcement learning), reflected by neural activity in the amygdala and midbrain. Furthermore, these reward-like learning signals are mirrored by opposite aversion-like signals in lateral orbitofrontal cortex and anterior cingulate cortex. This dual coding has parallels to 'opponent process' theories in psychology and promotes a formal account of prediction and expectation during pain.

Deep learning with convolutional neural network in radiology.

Science.gov (United States)

Yasaka, Koichiro; Akai, Hiroyuki; Kunimatsu, Akira; Kiryu, Shigeru; Abe, Osamu

2018-04-01

Deep learning with a convolutional neural network (CNN) is gaining attention recently for its high performance in image recognition. Images themselves can be utilized in a learning process with this technique, and feature extraction in advance of the learning process is not required. Important features can be automatically learned. Thanks to the development of hardware and software in addition to techniques regarding deep learning, application of this technique to radiological images for predicting clinically useful information, such as the detection and the evaluation of lesions, etc., are beginning to be investigated. This article illustrates basic technical knowledge regarding deep learning with CNNs along the actual course (collecting data, implementing CNNs, and training and testing phases). Pitfalls regarding this technique and how to manage them are also illustrated. We also described some advanced topics of deep learning, results of recent clinical studies, and the future directions of clinical application of deep learning techniques.

Radio frequency interference mitigation using deep convolutional neural networks

Science.gov (United States)

Akeret, J.; Chang, C.; Lucchi, A.; Refregier, A.

2017-01-01

We propose a novel approach for mitigating radio frequency interference (RFI) signals in radio data using the latest advances in deep learning. We employ a special type of Convolutional Neural Network, the U-Net, that enables the classification of clean signal and RFI signatures in 2D time-ordered data acquired from a radio telescope. We train and assess the performance of this network using the HIDE &SEEK radio data simulation and processing packages, as well as early Science Verification data acquired with the 7m single-dish telescope at the Bleien Observatory. We find that our U-Net implementation is showing competitive accuracy to classical RFI mitigation algorithms such as SEEK's SUMTHRESHOLD implementation. We publish our U-Net software package on GitHub under GPLv3 license.

Learning speaker-specific characteristics with a deep neural architecture.

Science.gov (United States)

Chen, Ke; Salman, Ahmad

2011-11-01

Speech signals convey various yet mixed information ranging from linguistic to speaker-specific information. However, most of acoustic representations characterize all different kinds of information as whole, which could hinder either a speech or a speaker recognition (SR) system from producing a better performance. In this paper, we propose a novel deep neural architecture (DNA) especially for learning speaker-specific characteristics from mel-frequency cepstral coefficients, an acoustic representation commonly used in both speech recognition and SR, which results in a speaker-specific overcomplete representation. In order to learn intrinsic speaker-specific characteristics, we come up with an objective function consisting of contrastive losses in terms of speaker similarity/dissimilarity and data reconstruction losses used as regularization to normalize the interference of non-speaker-related information. Moreover, we employ a hybrid learning strategy for learning parameters of the deep neural networks: i.e., local yet greedy layerwise unsupervised pretraining for initialization and global supervised learning for the ultimate discriminative goal. With four Linguistic Data Consortium (LDC) benchmarks and two non-English corpora, we demonstrate that our overcomplete representation is robust in characterizing various speakers, no matter whether their utterances have been used in training our DNA, and highly insensitive to text and languages spoken. Extensive comparative studies suggest that our approach yields favorite results in speaker verification and segmentation. Finally, we discuss several issues concerning our proposed approach.

Deep learning with convolutional neural networks: a resource for the control of robotic prosthetic hands via electromyography

Directory of Open Access Journals (Sweden)

Manfredo Atzori

2016-09-01

Full Text Available Motivation: Natural control methods based on surface electromyography and pattern recognition are promising for hand prosthetics. However, the control robustness offered by scientific research is still not sufficient for many real life applications and commercial prostheses are in the best case capable to offer natural control for only a few movements. Objective: In recent years deep learning revolutionized several fields of machine learning, including computer vision and speech recognition. Our objective is to test its capabilities for the natural control of robotic hands via surface electromyography by providing a baseline on a large number of intact and amputated subjects. Methods: We tested convolutional networks for the classification of an average of 50 hand movements in 67 intact subjects and 11 hand amputated subjects. The simple architecture of the neural network allowed to make several tests in order to evaluate the effect of pre-processing, layer architecture, data augmentation and optimization. The classification results are compared with a set of classical classification methods applied on the same datasets.Results: The classification accuracy obtained with convolutional neural networks using the proposed architecture is higher than the average results obtained with the classical classification methods but lower than the results obtained with the best reference methods in our tests. Significance: The results show that convolutional neural networks with a very simple architecture can produce accuracy comparable to the average classical classification methods. They show that several factors (including pre-processing, the architecture of the net and the optimization parameters can be fundamental for the analysis of surface electromyography data. Finally, the results suggest that deeper and more complex networks may increase dexterous control robustness, thus contributing to bridge the gap between the market and scientific research

Real-time cerebellar neuroprosthetic system based on a spiking neural network model of motor learning

Science.gov (United States)

Xu, Tao; Xiao, Na; Zhai, Xiaolong; Chan, Pak Kwan; Tin, Chung

2018-02-01

Objective. Damage to the brain, as a result of various medical conditions, impacts the everyday life of patients and there is still no complete cure to neurological disorders. Neuroprostheses that can functionally replace the damaged neural circuit have recently emerged as a possible solution to these problems. Here we describe the development of a real-time cerebellar neuroprosthetic system to substitute neural function in cerebellar circuitry for learning delay eyeblink conditioning (DEC). Approach. The system was empowered by a biologically realistic spiking neural network (SNN) model of the cerebellar neural circuit, which considers the neuronal population and anatomical connectivity of the network. The model simulated synaptic plasticity critical for learning DEC. This SNN model was carefully implemented on a field programmable gate array (FPGA) platform for real-time simulation. This hardware system was interfaced in in vivo experiments with anesthetized rats and it used neural spikes recorded online from the animal to learn and trigger conditioned eyeblink in the animal during training. Main results. This rat-FPGA hybrid system was able to process neuronal spikes in real-time with an embedded cerebellum model of ~10 000 neurons and reproduce learning of DEC with different inter-stimulus intervals. Our results validated that the system performance is physiologically relevant at both the neural (firing pattern) and behavioral (eyeblink pattern) levels. Significance. This integrated system provides the sufficient computation power for mimicking the cerebellar circuit in real-time. The system interacts with the biological system naturally at the spike level and can be generalized for including other neural components (neuron types and plasticity) and neural functions for potential neuroprosthetic applications.

Learning in Neural Networks: VLSI Implementation Strategies

Science.gov (United States)

Duong, Tuan Anh

1995-01-01

Fully-parallel hardware neural network implementations may be applied to high-speed recognition, classification, and mapping tasks in areas such as vision, or can be used as low-cost self-contained units for tasks such as error detection in mechanical systems (e.g. autos). Learning is required not only to satisfy application requirements, but also to overcome hardware-imposed limitations such as reduced dynamic range of connections.

Salient regions detection using convolutional neural networks and color volume

Science.gov (United States)

Liu, Guang-Hai; Hou, Yingkun

2018-03-01

Convolutional neural network is an important technique in machine learning, pattern recognition and image processing. In order to reduce the computational burden and extend the classical LeNet-5 model to the field of saliency detection, we propose a simple and novel computing model based on LeNet-5 network. In the proposed model, hue, saturation and intensity are utilized to extract depth cues, and then we integrate depth cues and color volume to saliency detection following the basic structure of the feature integration theory. Experimental results show that the proposed computing model outperforms some existing state-of-the-art methods on MSRA1000 and ECSSD datasets.

Image Classification, Deep Learning and Convolutional Neural Networks : A Comparative Study of Machine Learning Frameworks

OpenAIRE

Airola, Rasmus; Hager, Kristoffer

2017-01-01

The use of machine learning and specifically neural networks is a growing trend in software development, and has grown immensely in the last couple of years in the light of an increasing need to handle big data and large information flows. Machine learning has a broad area of application, such as human-computer interaction, predicting stock prices, real-time translation, and self driving vehicles. Large companies such as Microsoft and Google have already implemented machine learning in some o...

Continual and One-Shot Learning Through Neural Networks with Dynamic External Memory

DEFF Research Database (Denmark)

Lüders, Benno; Schläger, Mikkel; Korach, Aleksandra

2017-01-01

it easier to find unused memory location and therefor facilitates the evolution of continual learning networks. Our results suggest that augmenting evolving networks with an external memory component is not only a viable mechanism for adaptive behaviors in neuroevolution but also allows these networks...... a new task is learned. This paper takes a step in overcoming this limitation by building on the recently proposed Evolving Neural Turing Machine (ENTM) approach. In the ENTM, neural networks are augmented with an external memory component that they can write to and read from, which allows them to store...... associations quickly and over long periods of time. The results in this paper demonstrate that the ENTM is able to perform one-shot learning in reinforcement learning tasks without catastrophic forgetting of previously stored associations. Additionally, we introduce a new ENTM default jump mechanism that makes...

Neuromorphic implementations of neurobiological learning algorithms for spiking neural networks.

Science.gov (United States)

Walter, Florian; Röhrbein, Florian; Knoll, Alois

2015-12-01

The application of biologically inspired methods in design and control has a long tradition in robotics. Unlike previous approaches in this direction, the emerging field of neurorobotics not only mimics biological mechanisms at a relatively high level of abstraction but employs highly realistic simulations of actual biological nervous systems. Even today, carrying out these simulations efficiently at appropriate timescales is challenging. Neuromorphic chip designs specially tailored to this task therefore offer an interesting perspective for neurorobotics. Unlike Von Neumann CPUs, these chips cannot be simply programmed with a standard programming language. Like real brains, their functionality is determined by the structure of neural connectivity and synaptic efficacies. Enabling higher cognitive functions for neurorobotics consequently requires the application of neurobiological learning algorithms to adjust synaptic weights in a biologically plausible way. In this paper, we therefore investigate how to program neuromorphic chips by means of learning. First, we provide an overview over selected neuromorphic chip designs and analyze them in terms of neural computation, communication systems and software infrastructure. On the theoretical side, we review neurobiological learning techniques. Based on this overview, we then examine on-die implementations of these learning algorithms on the considered neuromorphic chips. A final discussion puts the findings of this work into context and highlights how neuromorphic hardware can potentially advance the field of autonomous robot systems. The paper thus gives an in-depth overview of neuromorphic implementations of basic mechanisms of synaptic plasticity which are required to realize advanced cognitive capabilities with spiking neural networks. Copyright © 2015 Elsevier Ltd. All rights reserved.

NetEnquiry--A Competitive Mobile Learning Approach for the Banking Sector

Science.gov (United States)

Beutner, Marc; Teine, Matthias; Gebbe, Marcel; Fortmann, Lara Melissa

2016-01-01

Initial and further education in the banking sector is becoming more and more important due to the fact that the regulations and the complexity in world of work and an international banking scene is increasing. In this article we provide the structures of and information on NetEnquiry, an innovative mobile learning environment in this field,…

The gamma model : a new neural network for temporal processing

NARCIS (Netherlands)

Vries, de B.

1992-01-01

In this paper we develop the gamma neural model, a new neural net architecture for processing of temporal patterns. Time varying patterns are normally segmented into a sequence of static patterns that are successively presented to a neural net. In the approach presented here segmentation is avoided.

Nonlinear signal processing using neural networks: Prediction and system modelling

Energy Technology Data Exchange (ETDEWEB)

Lapedes, A.; Farber, R.

1987-06-01

The backpropagation learning algorithm for neural networks is developed into a formalism for nonlinear signal processing. We illustrate the method by selecting two common topics in signal processing, prediction and system modelling, and show that nonlinear applications can be handled extremely well by using neural networks. The formalism is a natural, nonlinear extension of the linear Least Mean Squares algorithm commonly used in adaptive signal processing. Simulations are presented that document the additional performance achieved by using nonlinear neural networks. First, we demonstrate that the formalism may be used to predict points in a highly chaotic time series with orders of magnitude increase in accuracy over conventional methods including the Linear Predictive Method and the Gabor-Volterra-Weiner Polynomial Method. Deterministic chaos is thought to be involved in many physical situations including the onset of turbulence in fluids, chemical reactions and plasma physics. Secondly, we demonstrate the use of the formalism in nonlinear system modelling by providing a graphic example in which it is clear that the neural network has accurately modelled the nonlinear transfer function. It is interesting to note that the formalism provides explicit, analytic, global, approximations to the nonlinear maps underlying the various time series. Furthermore, the neural net seems to be extremely parsimonious in its requirements for data points from the time series. We show that the neural net is able to perform well because it globally approximates the relevant maps by performing a kind of generalized mode decomposition of the maps. 24 refs., 13 figs.

Biologically-inspired On-chip Learning in Pulsed Neural Networks

DEFF Research Database (Denmark)

Lehmann, Torsten; Woodburn, Robin

1999-01-01

Self-learning chips to implement many popular ANN (artificial neural network) algorithms are very difficult to design. We explain why this is so and say what lessons previous work teaches us in the design of self-learning systems. We offer a contribution to the "biologically-inspired" approach......, explaining what we mean by this term and providing an example of a robust, self-learning design that can solve simple classical-conditioning tasks, We give details of the design of individual circuits to perform component functions, which can then be combined into a network to solve the task. We argue...

Explaining neural signals in human visual cortex with an associative learning model.

Science.gov (United States)

Jiang, Jiefeng; Schmajuk, Nestor; Egner, Tobias

2012-08-01

"Predictive coding" models posit a key role for associative learning in visual cognition, viewing perceptual inference as a process of matching (learned) top-down predictions (or expectations) against bottom-up sensory evidence. At the neural level, these models propose that each region along the visual processing hierarchy entails one set of processing units encoding predictions of bottom-up input, and another set computing mismatches (prediction error or surprise) between predictions and evidence. This contrasts with traditional views of visual neurons operating purely as bottom-up feature detectors. In support of the predictive coding hypothesis, a recent human neuroimaging study (Egner, Monti, & Summerfield, 2010) showed that neural population responses to expected and unexpected face and house stimuli in the "fusiform face area" (FFA) could be well-described as a summation of hypothetical face-expectation and -surprise signals, but not by feature detector responses. Here, we used computer simulations to test whether these imaging data could be formally explained within the broader framework of a mathematical neural network model of associative learning (Schmajuk, Gray, & Lam, 1996). Results show that FFA responses could be fit very closely by model variables coding for conditional predictions (and their violations) of stimuli that unconditionally activate the FFA. These data document that neural population signals in the ventral visual stream that deviate from classic feature detection responses can formally be explained by associative prediction and surprise signals.

Using machine learning, neural networks and statistics to predict bankruptcy

NARCIS (Netherlands)

Pompe, P.P.M.; Feelders, A.J.; Feelders, A.J.

1997-01-01

Recent literature strongly suggests that machine learning approaches to classification outperform "classical" statistical methods. We make a comparison between the performance of linear discriminant analysis, classification trees, and neural networks in predicting corporate bankruptcy. Linear

Media Multitasking and Cognitive, Psychological, Neural, and Learning Differences.

Science.gov (United States)

Uncapher, Melina R; Lin, Lin; Rosen, Larry D; Kirkorian, Heather L; Baron, Naomi S; Bailey, Kira; Cantor, Joanne; Strayer, David L; Parsons, Thomas D; Wagner, Anthony D

2017-11-01

American youth spend more time with media than any other waking activity: an average of 7.5 hours per day, every day. On average, 29% of that time is spent juggling multiple media streams simultaneously (ie, media multitasking). This phenomenon is not limited to American youth but is paralleled across the globe. Given that a large number of media multitaskers (MMTs) are children and young adults whose brains are still developing, there is great urgency to understand the neurocognitive profiles of MMTs. It is critical to understand the relation between the relevant cognitive domains and underlying neural structure and function. Of equal importance is understanding the types of information processing that are necessary in 21st century learning environments. The present review surveys the growing body of evidence demonstrating that heavy MMTs show differences in cognition (eg, poorer memory), psychosocial behavior (eg, increased impulsivity), and neural structure (eg, reduced volume in anterior cingulate cortex). Furthermore, research indicates that multitasking with media during learning (in class or at home) can negatively affect academic outcomes. Until the direction of causality is understood (whether media multitasking causes such behavioral and neural differences or whether individuals with such differences tend to multitask with media more often), the data suggest that engagement with concurrent media streams should be thoughtfully considered. Findings from such research promise to inform policy and practice on an increasingly urgent societal issue while significantly advancing our understanding of the intersections between cognitive, psychosocial, neural, and academic factors. Copyright © 2017 by the American Academy of Pediatrics.

A neural learning classifier system with self-adaptive constructivism for mobile robot control.

Science.gov (United States)

Hurst, Jacob; Bull, Larry

2006-01-01

For artificial entities to achieve true autonomy and display complex lifelike behavior, they will need to exploit appropriate adaptable learning algorithms. In this context adaptability implies flexibility guided by the environment at any given time and an open-ended ability to learn appropriate behaviors. This article examines the use of constructivism-inspired mechanisms within a neural learning classifier system architecture that exploits parameter self-adaptation as an approach to realize such behavior. The system uses a rule structure in which each rule is represented by an artificial neural network. It is shown that appropriate internal rule complexity emerges during learning at a rate controlled by the learner and that the structure indicates underlying features of the task. Results are presented in simulated mazes before moving to a mobile robot platform.

ANT Advanced Neural Tool

Energy Technology Data Exchange (ETDEWEB)

Labrador, I.; Carrasco, R.; Martinez, L.

1996-07-01

This paper describes a practical introduction to the use of Artificial Neural Networks. Artificial Neural Nets are often used as an alternative to the traditional symbolic manipulation and first order logic used in Artificial Intelligence, due the high degree of difficulty to solve problems that can not be handled by programmers using algorithmic strategies. As a particular case of Neural Net a Multilayer Perception developed by programming in C language on OS9 real time operating system is presented. A detailed description about the program structure and practical use are included. Finally, several application examples that have been treated with the tool are presented, and some suggestions about hardware implementations. (Author) 15 refs.

ANT Advanced Neural Tool

International Nuclear Information System (INIS)

Labrador, I.; Carrasco, R.; Martinez, L.

1996-01-01

This paper describes a practical introduction to the use of Artificial Neural Networks. Artificial Neural Nets are often used as an alternative to the traditional symbolic manipulation and first order logic used in Artificial Intelligence, due the high degree of difficulty to solve problems that can not be handled by programmers using algorithmic strategies. As a particular case of Neural Net a Multilayer Perception developed by programming in C language on OS9 real time operating system is presented. A detailed description about the program structure and practical use are included. Finally, several application examples that have been treated with the tool are presented, and some suggestions about hardware implementations. (Author) 15 refs

Competitive Learning Neural Network Ensemble Weighted by Predicted Performance

Science.gov (United States)

Ye, Qiang

2010-01-01

Ensemble approaches have been shown to enhance classification by combining the outputs from a set of voting classifiers. Diversity in error patterns among base classifiers promotes ensemble performance. Multi-task learning is an important characteristic for Neural Network classifiers. Introducing a secondary output unit that receives different…

Supervised neural network modeling: an empirical investigation into learning from imbalanced data with labeling errors.

Science.gov (United States)

Khoshgoftaar, Taghi M; Van Hulse, Jason; Napolitano, Amri

2010-05-01

Neural network algorithms such as multilayer perceptrons (MLPs) and radial basis function networks (RBFNets) have been used to construct learners which exhibit strong predictive performance. Two data related issues that can have a detrimental impact on supervised learning initiatives are class imbalance and labeling errors (or class noise). Imbalanced data can make it more difficult for the neural network learning algorithms to distinguish between examples of the various classes, and class noise can lead to the formulation of incorrect hypotheses. Both class imbalance and labeling errors are pervasive problems encountered in a wide variety of application domains. Many studies have been performed to investigate these problems in isolation, but few have focused on their combined effects. This study presents a comprehensive empirical investigation using neural network algorithms to learn from imbalanced data with labeling errors. In particular, the first component of our study investigates the impact of class noise and class imbalance on two common neural network learning algorithms, while the second component considers the ability of data sampling (which is commonly used to address the issue of class imbalance) to improve their performances. Our results, for which over two million models were trained and evaluated, show that conclusions drawn using the more commonly studied C4.5 classifier may not apply when using neural networks.

Breast Cancer Diagnosis using Artificial Neural Networks with Extreme Learning Techniques

OpenAIRE

Chandra Prasetyo Utomo; Aan Kardiana; Rika Yuliwulandari

2014-01-01

Breast cancer is the second cause of dead among women. Early detection followed by appropriate cancer treatment can reduce the deadly risk. Medical professionals can make mistakes while identifying a disease. The help of technology such as data mining and machine learning can substantially improve the diagnosis accuracy. Artificial Neural Networks (ANN) has been widely used in intelligent breast cancer diagnosis. However, the standard Gradient-Based Back Propagation Artificial Neural Networks...

Using Deep Learning Neural Networks To Find Best Performing Audience Segments

Directory of Open Access Journals (Sweden)

Anup Badhe

2015-08-01

Full Text Available Finding the appropriate mobile audience for mobile advertising is always challenging since there are many data points that need to be considered and assimilated before a target segment can be created and used in ad serving by any ad server. Deep learning neural networks have been used in machine learning to use multiple processing layers to interpret large datasets with multiple dimensions to come up with a high-level characterization of the data. During a request for an advertisement and subsequently serving of the advertisement on the mobile device there are many trackers that are fired collecting a lot of data points. If the user likes the advertisement and clicks on it another set of trackers give additional information resulting from the click. This information is aggregated by the ad server and shown in its reporting console. The same information can form the basis of machine learning by feeding this information to a deep learning neural network to come up with audiences that can be targeted based on the product that is advertised.

Event-Driven Random Back-Propagation: Enabling Neuromorphic Deep Learning Machines

OpenAIRE

Neftci, Emre O.; Augustine, Charles; Paul, Somnath; Detorakis, Georgios

2017-01-01

An ongoing challenge in neuromorphic computing is to devise general and computationally efficient models of inference and learning which are compatible with the spatial and temporal constraints of the brain. One increasingly popular and successful approach is to take inspiration from inference and learning algorithms used in deep neural networks. However, the workhorse of deep learning, the gradient descent Gradient Back Propagation (BP) rule, often relies on the immediate availability of net...

Deep convolutional neural networks for automatic classification of gastric carcinoma using whole slide images in digital histopathology.

Science.gov (United States)

Sharma, Harshita; Zerbe, Norman; Klempert, Iris; Hellwich, Olaf; Hufnagl, Peter

2017-11-01

Deep learning using convolutional neural networks is an actively emerging field in histological image analysis. This study explores deep learning methods for computer-aided classification in H&E stained histopathological whole slide images of gastric carcinoma. An introductory convolutional neural network architecture is proposed for two computerized applications, namely, cancer classification based on immunohistochemical response and necrosis detection based on the existence of tumor necrosis in the tissue. Classification performance of the developed deep learning approach is quantitatively compared with traditional image analysis methods in digital histopathology requiring prior computation of handcrafted features, such as statistical measures using gray level co-occurrence matrix, Gabor filter-bank responses, LBP histograms, gray histograms, HSV histograms and RGB histograms, followed by random forest machine learning. Additionally, the widely known AlexNet deep convolutional framework is comparatively analyzed for the corresponding classification problems. The proposed convolutional neural network architecture reports favorable results, with an overall classification accuracy of 0.6990 for cancer classification and 0.8144 for necrosis detection. Copyright © 2017 Elsevier Ltd. All rights reserved.

Self-learning Monte Carlo with deep neural networks

Science.gov (United States)

Shen, Huitao; Liu, Junwei; Fu, Liang

2018-05-01

The self-learning Monte Carlo (SLMC) method is a general algorithm to speedup MC simulations. Its efficiency has been demonstrated in various systems by introducing an effective model to propose global moves in the configuration space. In this paper, we show that deep neural networks can be naturally incorporated into SLMC, and without any prior knowledge can learn the original model accurately and efficiently. Demonstrated in quantum impurity models, we reduce the complexity for a local update from O (β2) in Hirsch-Fye algorithm to O (β lnβ ) , which is a significant speedup especially for systems at low temperatures.

Minimal-Learning-Parameter Technique Based Adaptive Neural Sliding Mode Control of MEMS Gyroscope

Directory of Open Access Journals (Sweden)

Bin Xu

2017-01-01

Full Text Available This paper investigates an adaptive neural sliding mode controller for MEMS gyroscopes with minimal-learning-parameter technique. Considering the system uncertainty in dynamics, neural network is employed for approximation. Minimal-learning-parameter technique is constructed to decrease the number of update parameters, and in this way the computation burden is greatly reduced. Sliding mode control is designed to cancel the effect of time-varying disturbance. The closed-loop stability analysis is established via Lyapunov approach. Simulation results are presented to demonstrate the effectiveness of the method.

Learning to Recognize Actions From Limited Training Examples Using a Recurrent Spiking Neural Model

Science.gov (United States)

Panda, Priyadarshini; Srinivasa, Narayan

2018-01-01

A fundamental challenge in machine learning today is to build a model that can learn from few examples. Here, we describe a reservoir based spiking neural model for learning to recognize actions with a limited number of labeled videos. First, we propose a novel encoding, inspired by how microsaccades influence visual perception, to extract spike information from raw video data while preserving the temporal correlation across different frames. Using this encoding, we show that the reservoir generalizes its rich dynamical activity toward signature action/movements enabling it to learn from few training examples. We evaluate our approach on the UCF-101 dataset. Our experiments demonstrate that our proposed reservoir achieves 81.3/87% Top-1/Top-5 accuracy, respectively, on the 101-class data while requiring just 8 video examples per class for training. Our results establish a new benchmark for action recognition from limited video examples for spiking neural models while yielding competitive accuracy with respect to state-of-the-art non-spiking neural models. PMID:29551962

NetTurnP – Neural Network Prediction of Beta-turns by Use of Evolutionary Information and Predicted Protein Sequence Features

Science.gov (United States)

Petersen, Bent; Lundegaard, Claus; Petersen, Thomas Nordahl

2010-01-01

β-turns are the most common type of non-repetitive structures, and constitute on average 25% of the amino acids in proteins. The formation of β-turns plays an important role in protein folding, protein stability and molecular recognition processes. In this work we present the neural network method NetTurnP, for prediction of two-class β-turns and prediction of the individual β-turn types, by use of evolutionary information and predicted protein sequence features. It has been evaluated against a commonly used dataset BT426, and achieves a Matthews correlation coefficient of 0.50, which is the highest reported performance on a two-class prediction of β-turn and not-β-turn. Furthermore NetTurnP shows improved performance on some of the specific β-turn types. In the present work, neural network methods have been trained to predict β-turn or not and individual β-turn types from the primary amino acid sequence. The individual β-turn types I, I', II, II', VIII, VIa1, VIa2, VIba and IV have been predicted based on classifications by PROMOTIF, and the two-class prediction of β-turn or not is a superset comprised of all β-turn types. The performance is evaluated using a golden set of non-homologous sequences known as BT426. Our two-class prediction method achieves a performance of: MCC  = 0.50, Qtotal = 82.1%, sensitivity  = 75.6%, PPV  = 68.8% and AUC  = 0.864. We have compared our performance to eleven other prediction methods that obtain Matthews correlation coefficients in the range of 0.17 – 0.47. For the type specific β-turn predictions, only type I and II can be predicted with reasonable Matthews correlation coefficients, where we obtain performance values of 0.36 and 0.31, respectively. Conclusion The NetTurnP method has been implemented as a webserver, which is freely available at http://www.cbs.dtu.dk/services/NetTurnP/. NetTurnP is the only available webserver that allows submission of multiple sequences. PMID:21152409

Criticality meets learning: Criticality signatures in a self-organizing recurrent neural network.

Science.gov (United States)

Del Papa, Bruno; Priesemann, Viola; Triesch, Jochen

2017-01-01

Many experiments have suggested that the brain operates close to a critical state, based on signatures of criticality such as power-law distributed neuronal avalanches. In neural network models, criticality is a dynamical state that maximizes information processing capacities, e.g. sensitivity to input, dynamical range and storage capacity, which makes it a favorable candidate state for brain function. Although models that self-organize towards a critical state have been proposed, the relation between criticality signatures and learning is still unclear. Here, we investigate signatures of criticality in a self-organizing recurrent neural network (SORN). Investigating criticality in the SORN is of particular interest because it has not been developed to show criticality. Instead, the SORN has been shown to exhibit spatio-temporal pattern learning through a combination of neural plasticity mechanisms and it reproduces a number of biological findings on neural variability and the statistics and fluctuations of synaptic efficacies. We show that, after a transient, the SORN spontaneously self-organizes into a dynamical state that shows criticality signatures comparable to those found in experiments. The plasticity mechanisms are necessary to attain that dynamical state, but not to maintain it. Furthermore, onset of external input transiently changes the slope of the avalanche distributions - matching recent experimental findings. Interestingly, the membrane noise level necessary for the occurrence of the criticality signatures reduces the model's performance in simple learning tasks. Overall, our work shows that the biologically inspired plasticity and homeostasis mechanisms responsible for the SORN's spatio-temporal learning abilities can give rise to criticality signatures in its activity when driven by random input, but these break down under the structured input of short repeating sequences.

Deep Learning with Convolutional Neural Networks Applied to Electromyography Data: A Resource for the Classification of Movements for Prosthetic Hands.

Science.gov (United States)

Atzori, Manfredo; Cognolato, Matteo; Müller, Henning

2016-01-01

Natural control methods based on surface electromyography (sEMG) and pattern recognition are promising for hand prosthetics. However, the control robustness offered by scientific research is still not sufficient for many real life applications, and commercial prostheses are capable of offering natural control for only a few movements. In recent years deep learning revolutionized several fields of machine learning, including computer vision and speech recognition. Our objective is to test its methods for natural control of robotic hands via sEMG using a large number of intact subjects and amputees. We tested convolutional networks for the classification of an average of 50 hand movements in 67 intact subjects and 11 transradial amputees. The simple architecture of the neural network allowed to make several tests in order to evaluate the effect of pre-processing, layer architecture, data augmentation and optimization. The classification results are compared with a set of classical classification methods applied on the same datasets. The classification accuracy obtained with convolutional neural networks using the proposed architecture is higher than the average results obtained with the classical classification methods, but lower than the results obtained with the best reference methods in our tests. The results show that convolutional neural networks with a very simple architecture can produce accurate results comparable to the average classical classification methods. They show that several factors (including pre-processing, the architecture of the net and the optimization parameters) can be fundamental for the analysis of sEMG data. Larger networks can achieve higher accuracy on computer vision and object recognition tasks. This fact suggests that it may be interesting to evaluate if larger networks can increase sEMG classification accuracy too.

Deep Learning with Convolutional Neural Networks Applied to Electromyography Data: A Resource for the Classification of Movements for Prosthetic Hands

Science.gov (United States)

Atzori, Manfredo; Cognolato, Matteo; Müller, Henning

2016-01-01

Natural control methods based on surface electromyography (sEMG) and pattern recognition are promising for hand prosthetics. However, the control robustness offered by scientific research is still not sufficient for many real life applications, and commercial prostheses are capable of offering natural control for only a few movements. In recent years deep learning revolutionized several fields of machine learning, including computer vision and speech recognition. Our objective is to test its methods for natural control of robotic hands via sEMG using a large number of intact subjects and amputees. We tested convolutional networks for the classification of an average of 50 hand movements in 67 intact subjects and 11 transradial amputees. The simple architecture of the neural network allowed to make several tests in order to evaluate the effect of pre-processing, layer architecture, data augmentation and optimization. The classification results are compared with a set of classical classification methods applied on the same datasets. The classification accuracy obtained with convolutional neural networks using the proposed architecture is higher than the average results obtained with the classical classification methods, but lower than the results obtained with the best reference methods in our tests. The results show that convolutional neural networks with a very simple architecture can produce accurate results comparable to the average classical classification methods. They show that several factors (including pre-processing, the architecture of the net and the optimization parameters) can be fundamental for the analysis of sEMG data. Larger networks can achieve higher accuracy on computer vision and object recognition tasks. This fact suggests that it may be interesting to evaluate if larger networks can increase sEMG classification accuracy too. PMID:27656140

Identification of Complex Dynamical Systems with Neural Networks (2/2)

CERN Multimedia

CERN. Geneva

2016-01-01

The identification and analysis of high dimensional nonlinear systems is obviously a challenging task. Neural networks have been proven to be universal approximators but this still leaves the identification task a hard one. To do it efficiently, we have to violate some of the rules of classical regression theory. Furthermore we should focus on the interpretation of the resulting model to overcome its black box character. First, we will discuss function approximation with 3 layer feedforward neural networks up to new developments in deep neural networks and deep learning. These nets are not only of interest in connection with image analysis but are a center point of the current artificial intelligence developments. Second, we will focus on the analysis of complex dynamical system in the form of state space models realized as recurrent neural networks. After the introduction of small open dynamical systems we will study dynamical systems on manifolds. Here manifold and dynamics have to be identified in parall...

Identification of Complex Dynamical Systems with Neural Networks (1/2)

CERN Multimedia

CERN. Geneva

2016-01-01

The identification and analysis of high dimensional nonlinear systems is obviously a challenging task. Neural networks have been proven to be universal approximators but this still leaves the identification task a hard one. To do it efficiently, we have to violate some of the rules of classical regression theory. Furthermore we should focus on the interpretation of the resulting model to overcome its black box character. First, we will discuss function approximation with 3 layer feedforward neural networks up to new developments in deep neural networks and deep learning. These nets are not only of interest in connection with image analysis but are a center point of the current artificial intelligence developments. Second, we will focus on the analysis of complex dynamical system in the form of state space models realized as recurrent neural networks. After the introduction of small open dynamical systems we will study dynamical systems on manifolds. Here manifold and dynamics have to be identified in parall...

Differential theory of learning for efficient neural network pattern recognition

Science.gov (United States)

Hampshire, John B., II; Vijaya Kumar, Bhagavatula

1993-09-01

We describe a new theory of differential learning by which a broad family of pattern classifiers (including many well-known neural network paradigms) can learn stochastic concepts efficiently. We describe the relationship between a classifier's ability to generate well to unseen test examples and the efficiency of the strategy by which it learns. We list a series of proofs that differential learning is efficient in its information and computational resource requirements, whereas traditional probabilistic learning strategies are not. The proofs are illustrated by a simple example that lends itself to closed-form analysis. We conclude with an optical character recognition task for which three different types of differentially generated classifiers generalize significantly better than their probabilistically generated counterparts.

Outsmarting neural networks: an alternative paradigm for machine learning

Energy Technology Data Exchange (ETDEWEB)

Protopopescu, V.; Rao, N.S.V.

1996-10-01

We address three problems in machine learning, namely: (i) function learning, (ii) regression estimation, and (iii) sensor fusion, in the Probably and Approximately Correct (PAC) framework. We show that, under certain conditions, one can reduce the three problems above to the regression estimation. The latter is usually tackled with artificial neural networks (ANNs) that satisfy the PAC criteria, but have high computational complexity. We propose several computationally efficient PAC alternatives to ANNs to solve the regression estimation. Thereby we also provide efficient PAC solutions to the function learning and sensor fusion problems. The approach is based on cross-fertilizing concepts and methods from statistical estimation, nonlinear algorithms, and the theory of computational complexity, and is designed as part of a new, coherent paradigm for machine learning.

Design and FPGA-implementation of multilayer neural networks with on-chip learning

International Nuclear Information System (INIS)

Haggag, S.S.M.Y

2008-01-01

Artificial Neural Networks (ANN) is used in many applications in the industry because of their parallel structure, high speed, and their ability to give easy solution to complicated problems. For example identifying the orange and apple in the sorting machine with neural network is easier than using image processing techniques to do the same thing. There are different software for designing, training, and testing the ANN, but in order to use the ANN in the industry, it should be implemented on hardware outside the computer. Neural networks are artificial systems inspired on the brain's cognitive behavior, which can learn tasks with some degree of complexity, such as signal processing, diagnosis, robotics, image processing, and pattern recognition. Many applications demand a high computing power and the traditional software implementation are not sufficient.This thesis presents design and FPGA implementation of Multilayer Neural Networks with On-chip learning in re-configurable hardware. Hardware implementation of neural network algorithm is very interesting due their high performance and they can easily be made parallel. The architecture proposed herein takes advantage of distinct data paths for the forward and backward propagation stages and a pipelined adaptation of the on- line backpropagation algorithm to significantly improve the performance of the learning phase. The architecture is easily scalable and able to cope with arbitrary network sizes with the same hardware. The implementation is targeted diagnosis of the Research Reactor accidents to avoid the risk of occurrence of a nuclear accident. The proposed designed circuits are implemented using Xilinx FPGA Chip XC40150xv and occupied 73% of Chip CLBs. It achieved 10.8 μs to take decision in the forward propagation compared with current software implemented of RPS which take 24 ms. The results show that the proposed architecture leads to significant speed up comparing to high end software solutions. On

Larger bases and mixed analog/digital neural nets

Energy Technology Data Exchange (ETDEWEB)

Beiu, V.

1998-12-31

The paper overviews results dealing with the approximation capabilities of neural networks, and bounds on the size of threshold gate circuits. Based on an explicit numerical algorithm for Kolmogorov`s superpositions the authors show that minimum size neural networks--for implementing any Boolean function--have the identity function as the activation function. Conclusions and several comments on the required precision are ending the paper.

Neural correlates of face gender discrimination learning.

Science.gov (United States)

Su, Junzhu; Tan, Qingleng; Fang, Fang

2013-04-01

Using combined psychophysics and event-related potentials (ERPs), we investigated the effect of perceptual learning on face gender discrimination and probe the neural correlates of the learning effect. Human subjects were trained to perform a gender discrimination task with male or female faces. Before and after training, they were tested with the trained faces and other faces with the same and opposite genders. ERPs responding to these faces were recorded. Psychophysical results showed that training significantly improved subjects' discrimination performance and the improvement was specific to the trained gender, as well as to the trained identities. The training effect indicates that learning occurs at two levels-the category level (gender) and the exemplar level (identity). ERP analyses showed that the gender and identity learning was associated with the N170 latency reduction at the left occipital-temporal area and the N170 amplitude reduction at the right occipital-temporal area, respectively. These findings provide evidence for the facilitation model and the sharpening model on neuronal plasticity from visual experience, suggesting a faster processing speed and a sparser representation of face induced by perceptual learning.

Neural Basis of Reinforcement Learning and Decision Making

Science.gov (United States)

Lee, Daeyeol; Seo, Hyojung; Jung, Min Whan

2012-01-01

Reinforcement learning is an adaptive process in which an animal utilizes its previous experience to improve the outcomes of future choices. Computational theories of reinforcement learning play a central role in the newly emerging areas of neuroeconomics and decision neuroscience. In this framework, actions are chosen according to their value functions, which describe how much future reward is expected from each action. Value functions can be adjusted not only through reward and penalty, but also by the animal’s knowledge of its current environment. Studies have revealed that a large proportion of the brain is involved in representing and updating value functions and using them to choose an action. However, how the nature of a behavioral task affects the neural mechanisms of reinforcement learning remains incompletely understood. Future studies should uncover the principles by which different computational elements of reinforcement learning are dynamically coordinated across the entire brain. PMID:22462543

2D neural hardware versus 3D biological ones

Energy Technology Data Exchange (ETDEWEB)

Beiu, V.

1998-12-31

This paper will present important limitations of hardware neural nets as opposed to biological neural nets (i.e. the real ones). The author starts by discussing neural structures and their biological inspirations, while mentioning the simplifications leading to artificial neural nets. Going further, the focus will be on hardware constraints. The author will present recent results for three different alternatives of implementing neural networks: digital, threshold gate, and analog, while the area and the delay will be related to neurons' fan-in and weights' precision. Based on all of these, it will be shown why hardware implementations cannot cope with their biological inspiration with respect to their power of computation: the mapping onto silicon lacking the third dimension of biological nets. This translates into reduced fan-in, and leads to reduced precision. The main conclusion is that one is faced with the following alternatives: (1) try to cope with the limitations imposed by silicon, by speeding up the computation of the elementary silicon neurons; (2) investigate solutions which would allow one to use the third dimension, e.g. using optical interconnections.

Classification of breast cancer cytological specimen using convolutional neural network

Science.gov (United States)

Żejmo, Michał; Kowal, Marek; Korbicz, Józef; Monczak, Roman

2017-01-01

The paper presents a deep learning approach for automatic classification of breast tumors based on fine needle cytology. The main aim of the system is to distinguish benign from malignant cases based on microscopic images. Experiment was carried out on cytological samples derived from 50 patients (25 benign cases + 25 malignant cases) diagnosed in Regional Hospital in Zielona Góra. To classify microscopic images, we used convolutional neural networks (CNN) of two types: GoogLeNet and AlexNet. Due to the very large size of images of cytological specimen (on average 200000 × 100000 pixels), they were divided into smaller patches of size 256 × 256 pixels. Breast cancer classification usually is based on morphometric features of nuclei. Therefore, training and validation patches were selected using Support Vector Machine (SVM) so that suitable amount of cell material was depicted. Neural classifiers were tuned using GPU accelerated implementation of gradient descent algorithm. Training error was defined as a cross-entropy classification loss. Classification accuracy was defined as the percentage ratio of successfully classified validation patches to the total number of validation patches. The best accuracy rate of 83% was obtained by GoogLeNet model. We observed that more misclassified patches belong to malignant cases.

Neural correlates of threat perception: neural equivalence of conspecific and heterospecific mobbing calls is learned.

Directory of Open Access Journals (Sweden)

Marc T Avey

Full Text Available Songbird auditory areas (i.e., CMM and NCM are preferentially activated to playback of conspecific vocalizations relative to heterospecific and arbitrary noise. Here, we asked if the neural response to auditory stimulation is not simply preferential for conspecific vocalizations but also for the information conveyed by the vocalization. Black-capped chickadees use their chick-a-dee mobbing call to recruit conspecifics and other avian species to mob perched predators. Mobbing calls produced in response to smaller, higher-threat predators contain more "D" notes compared to those produced in response to larger, lower-threat predators and thus convey the degree of threat of predators. We specifically asked whether the neural response varies with the degree of threat conveyed by the mobbing calls of chickadees and whether the neural response is the same for actual predator calls that correspond to the degree of threat of the chickadee mobbing calls. Our results demonstrate that, as degree of threat increases in conspecific chickadee mobbing calls, there is a corresponding increase in immediate early gene (IEG expression in telencephalic auditory areas. We also demonstrate that as the degree of threat increases for the heterospecific predator, there is a corresponding increase in IEG expression in the auditory areas. Furthermore, there was no significant difference in the amount IEG expression between conspecific mobbing calls or heterospecific predator calls that were the same degree of threat. In a second experiment, using hand-reared chickadees without predator experience, we found more IEG expression in response to mobbing calls than corresponding predator calls, indicating that degree of threat is learned. Our results demonstrate that degree of threat corresponds to neural activity in the auditory areas and that threat can be conveyed by different species signals and that these signals must be learned.

Neural Correlates of Threat Perception: Neural Equivalence of Conspecific and Heterospecific Mobbing Calls Is Learned

Science.gov (United States)

Avey, Marc T.; Hoeschele, Marisa; Moscicki, Michele K.; Bloomfield, Laurie L.; Sturdy, Christopher B.

2011-01-01

Songbird auditory areas (i.e., CMM and NCM) are preferentially activated to playback of conspecific vocalizations relative to heterospecific and arbitrary noise [1]–[2]. Here, we asked if the neural response to auditory stimulation is not simply preferential for conspecific vocalizations but also for the information conveyed by the vocalization. Black-capped chickadees use their chick-a-dee mobbing call to recruit conspecifics and other avian species to mob perched predators [3]. Mobbing calls produced in response to smaller, higher-threat predators contain more “D” notes compared to those produced in response to larger, lower-threat predators and thus convey the degree of threat of predators [4]. We specifically asked whether the neural response varies with the degree of threat conveyed by the mobbing calls of chickadees and whether the neural response is the same for actual predator calls that correspond to the degree of threat of the chickadee mobbing calls. Our results demonstrate that, as degree of threat increases in conspecific chickadee mobbing calls, there is a corresponding increase in immediate early gene (IEG) expression in telencephalic auditory areas. We also demonstrate that as the degree of threat increases for the heterospecific predator, there is a corresponding increase in IEG expression in the auditory areas. Furthermore, there was no significant difference in the amount IEG expression between conspecific mobbing calls or heterospecific predator calls that were the same degree of threat. In a second experiment, using hand-reared chickadees without predator experience, we found more IEG expression in response to mobbing calls than corresponding predator calls, indicating that degree of threat is learned. Our results demonstrate that degree of threat corresponds to neural activity in the auditory areas and that threat can be conveyed by different species signals and that these signals must be learned. PMID:21909363

Neural correlates of threat perception: neural equivalence of conspecific and heterospecific mobbing calls is learned.

Science.gov (United States)

Avey, Marc T; Hoeschele, Marisa; Moscicki, Michele K; Bloomfield, Laurie L; Sturdy, Christopher B

2011-01-01

Songbird auditory areas (i.e., CMM and NCM) are preferentially activated to playback of conspecific vocalizations relative to heterospecific and arbitrary noise. Here, we asked if the neural response to auditory stimulation is not simply preferential for conspecific vocalizations but also for the information conveyed by the vocalization. Black-capped chickadees use their chick-a-dee mobbing call to recruit conspecifics and other avian species to mob perched predators. Mobbing calls produced in response to smaller, higher-threat predators contain more "D" notes compared to those produced in response to larger, lower-threat predators and thus convey the degree of threat of predators. We specifically asked whether the neural response varies with the degree of threat conveyed by the mobbing calls of chickadees and whether the neural response is the same for actual predator calls that correspond to the degree of threat of the chickadee mobbing calls. Our results demonstrate that, as degree of threat increases in conspecific chickadee mobbing calls, there is a corresponding increase in immediate early gene (IEG) expression in telencephalic auditory areas. We also demonstrate that as the degree of threat increases for the heterospecific predator, there is a corresponding increase in IEG expression in the auditory areas. Furthermore, there was no significant difference in the amount IEG expression between conspecific mobbing calls or heterospecific predator calls that were the same degree of threat. In a second experiment, using hand-reared chickadees without predator experience, we found more IEG expression in response to mobbing calls than corresponding predator calls, indicating that degree of threat is learned. Our results demonstrate that degree of threat corresponds to neural activity in the auditory areas and that threat can be conveyed by different species signals and that these signals must be learned.

Single-Iteration Learning Algorithm for Feed-Forward Neural Networks

Energy Technology Data Exchange (ETDEWEB)

Barhen, J.; Cogswell, R.; Protopopescu, V.

1999-07-31

A new methodology for neural learning is presented, whereby only a single iteration is required to train a feed-forward network with near-optimal results. To this aim, a virtual input layer is added to the multi-layer architecture. The virtual input layer is connected to the nominal input layer by a specird nonlinear transfer function, and to the fwst hidden layer by regular (linear) synapses. A sequence of alternating direction singular vrdue decompositions is then used to determine precisely the inter-layer synaptic weights. This algorithm exploits the known separability of the linear (inter-layer propagation) and nonlinear (neuron activation) aspects of information &ansfer within a neural network.

On the relationships between generative encodings, regularity, and learning abilities when evolving plastic artificial neural networks.

Directory of Open Access Journals (Sweden)

Paul Tonelli

Full Text Available A major goal of bio-inspired artificial intelligence is to design artificial neural networks with abilities that resemble those of animal nervous systems. It is commonly believed that two keys for evolving nature-like artificial neural networks are (1 the developmental process that links genes to nervous systems, which enables the evolution of large, regular neural networks, and (2 synaptic plasticity, which allows neural networks to change during their lifetime. So far, these two topics have been mainly studied separately. The present paper shows that they are actually deeply connected. Using a simple operant conditioning task and a classic evolutionary algorithm, we compare three ways to encode plastic neural networks: a direct encoding, a developmental encoding inspired by computational neuroscience models, and a developmental encoding inspired by morphogen gradients (similar to HyperNEAT. Our results suggest that using a developmental encoding could improve the learning abilities of evolved, plastic neural networks. Complementary experiments reveal that this result is likely the consequence of the bias of developmental encodings towards regular structures: (1 in our experimental setup, encodings that tend to produce more regular networks yield networks with better general learning abilities; (2 whatever the encoding is, networks that are the more regular are statistically those that have the best learning abilities.

Learning behavior and temporary minima of two-layer neural networks

NARCIS (Netherlands)

Annema, Anne J.; Hoen, Klaas; Hoen, Klaas; Wallinga, Hans

1994-01-01

This paper presents a mathematical analysis of the occurrence of temporary minima during training of a single-output, two-layer neural network, with learning according to the back-propagation algorithm. A new vector decomposition method is introduced, which simplifies the mathematical analysis of

NutriNet: A Deep Learning Food and Drink Image Recognition System for Dietary Assessment.

Science.gov (United States)

Mezgec, Simon; Koroušić Seljak, Barbara

2017-06-27

Automatic food image recognition systems are alleviating the process of food-intake estimation and dietary assessment. However, due to the nature of food images, their recognition is a particularly challenging task, which is why traditional approaches in the field have achieved a low classification accuracy. Deep neural networks have outperformed such solutions, and we present a novel approach to the problem of food and drink image detection and recognition that uses a newly-defined deep convolutional neural network architecture, called NutriNet. This architecture was tuned on a recognition dataset containing 225,953 512 × 512 pixel images of 520 different food and drink items from a broad spectrum of food groups, on which we achieved a classification accuracy of 86 . 72 % , along with an accuracy of 94 . 47 % on a detection dataset containing 130 , 517 images. We also performed a real-world test on a dataset of self-acquired images, combined with images from Parkinson's disease patients, all taken using a smartphone camera, achieving a top-five accuracy of 55 % , which is an encouraging result for real-world images. Additionally, we tested NutriNet on the University of Milano-Bicocca 2016 (UNIMIB2016) food image dataset, on which we improved upon the provided baseline recognition result. An online training component was implemented to continually fine-tune the food and drink recognition model on new images. The model is being used in practice as part of a mobile app for the dietary assessment of Parkinson's disease patients.

Functionally segregated neural substrates for arbitrary audiovisual paired-association learning.

Science.gov (United States)

Tanabe, Hiroki C; Honda, Manabu; Sadato, Norihiro

2005-07-06

To clarify the neural substrates and their dynamics during crossmodal association learning, we conducted functional magnetic resonance imaging (MRI) during audiovisual paired-association learning of delayed matching-to-sample tasks. Thirty subjects were involved in the study; 15 performed an audiovisual paired-association learning task, and the remainder completed a control visuo-visual task. Each trial consisted of the successive presentation of a pair of stimuli. Subjects were asked to identify predefined audiovisual or visuo-visual pairs by trial and error. Feedback for each trial was given regardless of whether the response was correct or incorrect. During the delay period, several areas showed an increase in the MRI signal as learning proceeded: crossmodal activity increased in unimodal areas corresponding to visual or auditory areas, and polymodal responses increased in the occipitotemporal junction and parahippocampal gyrus. This pattern was not observed in the visuo-visual intramodal paired-association learning task, suggesting that crossmodal associations might be formed by binding unimodal sensory areas via polymodal regions. In both the audiovisual and visuo-visual tasks, the MRI signal in the superior temporal sulcus (STS) in response to the second stimulus and feedback peaked during the early phase of learning and then decreased, indicating that the STS might be key to the creation of paired associations, regardless of stimulus type. In contrast to the activity changes in the regions discussed above, there was constant activity in the frontoparietal circuit during the delay period in both tasks, implying that the neural substrates for the formation and storage of paired associates are distinct from working memory circuits.

Multi-categorical deep learning neural network to classify retinal images: A pilot study employing small database.

Science.gov (United States)

Choi, Joon Yul; Yoo, Tae Keun; Seo, Jeong Gi; Kwak, Jiyong; Um, Terry Taewoong; Rim, Tyler Hyungtaek

2017-01-01

Deep learning emerges as a powerful tool for analyzing medical images. Retinal disease detection by using computer-aided diagnosis from fundus image has emerged as a new method. We applied deep learning convolutional neural network by using MatConvNet for an automated detection of multiple retinal diseases with fundus photographs involved in STructured Analysis of the REtina (STARE) database. Dataset was built by expanding data on 10 categories, including normal retina and nine retinal diseases. The optimal outcomes were acquired by using a random forest transfer learning based on VGG-19 architecture. The classification results depended greatly on the number of categories. As the number of categories increased, the performance of deep learning models was diminished. When all 10 categories were included, we obtained results with an accuracy of 30.5%, relative classifier information (RCI) of 0.052, and Cohen's kappa of 0.224. Considering three integrated normal, background diabetic retinopathy, and dry age-related macular degeneration, the multi-categorical classifier showed accuracy of 72.8%, 0.283 RCI, and 0.577 kappa. In addition, several ensemble classifiers enhanced the multi-categorical classification performance. The transfer learning incorporated with ensemble classifier of clustering and voting approach presented the best performance with accuracy of 36.7%, 0.053 RCI, and 0.225 kappa in the 10 retinal diseases classification problem. First, due to the small size of datasets, the deep learning techniques in this study were ineffective to be applied in clinics where numerous patients suffering from various types of retinal disorders visit for diagnosis and treatment. Second, we found that the transfer learning incorporated with ensemble classifiers can improve the classification performance in order to detect multi-categorical retinal diseases. Further studies should confirm the effectiveness of algorithms with large datasets obtained from hospitals.

LEARNING ALGORITHM EFFECT ON MULTILAYER FEED FORWARD ARTIFICIAL NEURAL NETWORK PERFORMANCE IN IMAGE CODING

Directory of Open Access Journals (Sweden)

OMER MAHMOUD

2007-08-01

Full Text Available One of the essential factors that affect the performance of Artificial Neural Networks is the learning algorithm. The performance of Multilayer Feed Forward Artificial Neural Network performance in image compression using different learning algorithms is examined in this paper. Based on Gradient Descent, Conjugate Gradient, Quasi-Newton techniques three different error back propagation algorithms have been developed for use in training two types of neural networks, a single hidden layer network and three hidden layers network. The essence of this study is to investigate the most efficient and effective training methods for use in image compression and its subsequent applications. The obtained results show that the Quasi-Newton based algorithm has better performance as compared to the other two algorithms.

A Multiobjective Sparse Feature Learning Model for Deep Neural Networks.

Science.gov (United States)

Gong, Maoguo; Liu, Jia; Li, Hao; Cai, Qing; Su, Linzhi

2015-12-01

Hierarchical deep neural networks are currently popular learning models for imitating the hierarchical architecture of human brain. Single-layer feature extractors are the bricks to build deep networks. Sparse feature learning models are popular models that can learn useful representations. But most of those models need a user-defined constant to control the sparsity of representations. In this paper, we propose a multiobjective sparse feature learning model based on the autoencoder. The parameters of the model are learnt by optimizing two objectives, reconstruction error and the sparsity of hidden units simultaneously to find a reasonable compromise between them automatically. We design a multiobjective induced learning procedure for this model based on a multiobjective evolutionary algorithm. In the experiments, we demonstrate that the learning procedure is effective, and the proposed multiobjective model can learn useful sparse features.

The Role of Architectural and Learning Constraints in Neural Network Models: A Case Study on Visual Space Coding.

Science.gov (United States)

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

2017-01-01

The recent "deep learning revolution" in artificial neural networks had strong impact and widespread deployment for engineering applications, but the use of deep learning for neurocomputational modeling has been so far limited. In this article we argue that unsupervised deep learning represents an important step forward for improving neurocomputational models of perception and cognition, because it emphasizes the role of generative learning as opposed to discriminative (supervised) learning. As a case study, we present a series of simulations investigating the emergence of neural coding of visual space for sensorimotor transformations. We compare different network architectures commonly used as building blocks for unsupervised deep learning by systematically testing the type of receptive fields and gain modulation developed by the hidden neurons. In particular, we compare Restricted Boltzmann Machines (RBMs), which are stochastic, generative networks with bidirectional connections trained using contrastive divergence, with autoencoders, which are deterministic networks trained using error backpropagation. For both learning architectures we also explore the role of sparse coding, which has been identified as a fundamental principle of neural computation. The unsupervised models are then compared with supervised, feed-forward networks that learn an explicit mapping between different spatial reference frames. Our simulations show that both architectural and learning constraints strongly influenced the emergent coding of visual space in terms of distribution of tuning functions at the level of single neurons. Unsupervised models, and particularly RBMs, were found to more closely adhere to neurophysiological data from single-cell recordings in the primate parietal cortex. These results provide new insights into how basic properties of artificial neural networks might be relevant for modeling neural information processing in biological systems.

Neural mechanisms of human perceptual learning: electrophysiological evidence for a two-stage process.

Science.gov (United States)

Hamamé, Carlos M; Cosmelli, Diego; Henriquez, Rodrigo; Aboitiz, Francisco

2011-04-26

Humans and other animals change the way they perceive the world due to experience. This process has been labeled as perceptual learning, and implies that adult nervous systems can adaptively modify the way in which they process sensory stimulation. However, the mechanisms by which the brain modifies this capacity have not been sufficiently analyzed. We studied the neural mechanisms of human perceptual learning by combining electroencephalographic (EEG) recordings of brain activity and the assessment of psychophysical performance during training in a visual search task. All participants improved their perceptual performance as reflected by an increase in sensitivity (d') and a decrease in reaction time. The EEG signal was acquired throughout the entire experiment revealing amplitude increments, specific and unspecific to the trained stimulus, in event-related potential (ERP) components N2pc and P3 respectively. P3 unspecific modification can be related to context or task-based learning, while N2pc may be reflecting a more specific attentional-related boosting of target detection. Moreover, bell and U-shaped profiles of oscillatory brain activity in gamma (30-60 Hz) and alpha (8-14 Hz) frequency bands may suggest the existence of two phases for learning acquisition, which can be understood as distinctive optimization mechanisms in stimulus processing. We conclude that there are reorganizations in several neural processes that contribute differently to perceptual learning in a visual search task. We propose an integrative model of neural activity reorganization, whereby perceptual learning takes place as a two-stage phenomenon including perceptual, attentional and contextual processes.

Pro visual C++/CLI and the net 35 platform

CERN Document Server

Fraser, Stephen

2008-01-01

Pro Visual C++/CLI and the .NET 3.5 Platform is about writing .NET applications using C++/CLI. While readers are learning the ins and outs of .NET application development, they will also be learning the syntax of C++, both old and new to .NET. Readers will also gain a good understanding of the .NET architecture. This is truly a .NET book applying C++ as its development language not another C++ syntax book that happens to cover .NET.

Cooperate to Validate. Observal-Net Experts' Report on Validation of Non-Formal and Informal Learning (VNIL) 2013

Science.gov (United States)

Weber Guisan, Saskia; Voit, Janine; Lengauer, Sonja; Proinger, Eva; Duvekot, Ruud; Aagaard, Kirsten

2014-01-01

The present publication is one of the outcomes of the OBSERVAL-NET project (followup of the OBSERVAL project). The main aim of OBSERVAL-NET was to set up a stakeholder centric network of organisations supporting the validation of non-formal and informal learning in Europe based on the formation of national working groups in the 8 participating…

Cooperate to Validate: OBSERVAL-NET Experts' Report on Validation of Non-Formal and Informal Learning (VNIL) 2013

Science.gov (United States)

Weber Guisan, Saskia; Voit, Janine; Lengauer, Sonja; Proinger, Eva; Duvekot, Ruud; Aagaard, Kirsten

2014-01-01

The present publication is one of the outcomes of the OBSERVAL-NET project (follow-up of the OBSERVAL project). The main aim of OBSERVAL-NET was to set up a stakeholder-centric network of organisations supporting the validation of non-formal and informal learning in Europe based on the formation of national working groups in the 8 participating…

Learning representations for the early detection of sepsis with deep neural networks.

Science.gov (United States)

Kam, Hye Jin; Kim, Ha Young

2017-10-01

Sepsis is one of the leading causes of death in intensive care unit patients. Early detection of sepsis is vital because mortality increases as the sepsis stage worsens. This study aimed to develop detection models for the early stage of sepsis using deep learning methodologies, and to compare the feasibility and performance of the new deep learning methodology with those of the regression method with conventional temporal feature extraction. Study group selection adhered to the InSight model. The results of the deep learning-based models and the InSight model were compared. With deep feedforward networks, the area under the ROC curve (AUC) of the models were 0.887 and 0.915 for the InSight and the new feature sets, respectively. For the model with the combined feature set, the AUC was the same as that of the basic feature set (0.915). For the long short-term memory model, only the basic feature set was applied and the AUC improved to 0.929 compared with the existing 0.887 of the InSight model. The contributions of this paper can be summarized in three ways: (i) improved performance without feature extraction using domain knowledge, (ii) verification of feature extraction capability of deep neural networks through comparison with reference features, and (iii) improved performance with feedforward neural networks using long short-term memory, a neural network architecture that can learn sequential patterns. Copyright © 2017 Elsevier Ltd. All rights reserved.

Reconstruction of neutron spectra through neural networks

International Nuclear Information System (INIS)

Vega C, H.R.; Hernandez D, V.M.; Manzanares A, E.

2003-01-01

A neural network has been used to reconstruct the neutron spectra starting from the counting rates of the detectors of the Bonner sphere spectrophotometric system. A group of 56 neutron spectra was selected to calculate the counting rates that would produce in a Bonner sphere system, with these data and the spectra it was trained the neural network. To prove the performance of the net, 12 spectra were used, 6 were taken of the group used for the training, 3 were obtained of mathematical functions and those other 3 correspond to real spectra. When comparing the original spectra of those reconstructed by the net we find that our net has a poor performance when reconstructing monoenergetic spectra, this attributes it to those characteristic of the spectra used for the training of the neural network, however for the other groups of spectra the results of the net are appropriate with the prospective ones. (Author)

Culture in the mind's mirror: how anthropology and neuroscience can inform a model of the neural substrate for cultural imitative learning.

Science.gov (United States)

Losin, Elizabeth A Reynolds; Dapretto, Mirella; Iacoboni, Marco

2009-01-01

Cultural neuroscience, the study of how cultural experience shapes the brain, is an emerging subdiscipline in the neurosciences. Yet, a foundational question to the study of culture and the brain remains neglected by neuroscientific inquiry: "How does cultural information get into the brain in the first place?" Fortunately, the tools needed to explore the neural architecture of cultural learning - anthropological theories and cognitive neuroscience methodologies - already exist; they are merely separated by disciplinary boundaries. Here we review anthropological theories of cultural learning derived from fieldwork and modeling; since cultural learning theory suggests that sophisticated imitation abilities are at the core of human cultural learning, we focus our review on cultural imitative learning. Accordingly we proceed to discuss the neural underpinnings of imitation and other mechanisms important for cultural learning: learning biases, mental state attribution, and reinforcement learning. Using cultural neuroscience theory and cognitive neuroscience research as our guides, we then propose a preliminary model of the neural architecture of cultural learning. Finally, we discuss future studies needed to test this model and fully explore and explain the neural underpinnings of cultural imitative learning.

Selected Flight Test Results for Online Learning Neural Network-Based Flight Control System

Science.gov (United States)

Williams-Hayes, Peggy S.

2004-01-01

The NASA F-15 Intelligent Flight Control System project team developed a series of flight control concepts designed to demonstrate neural network-based adaptive controller benefits, with the objective to develop and flight-test control systems using neural network technology to optimize aircraft performance under nominal conditions and stabilize the aircraft under failure conditions. This report presents flight-test results for an adaptive controller using stability and control derivative values from an online learning neural network. A dynamic cell structure neural network is used in conjunction with a real-time parameter identification algorithm to estimate aerodynamic stability and control derivative increments to baseline aerodynamic derivatives in flight. This open-loop flight test set was performed in preparation for a future phase in which the learning neural network and parameter identification algorithm output would provide the flight controller with aerodynamic stability and control derivative updates in near real time. Two flight maneuvers are analyzed - pitch frequency sweep and automated flight-test maneuver designed to optimally excite the parameter identification algorithm in all axes. Frequency responses generated from flight data are compared to those obtained from nonlinear simulation runs. Flight data examination shows that addition of flight-identified aerodynamic derivative increments into the simulation improved aircraft pitch handling qualities.

Picasso: A Modular Framework for Visualizing the Learning Process of Neural Network Image Classifiers

Directory of Open Access Journals (Sweden)

Ryan Henderson

2017-09-01

Full Text Available Picasso is a free open-source (Eclipse Public License web application written in Python for rendering standard visualizations useful for analyzing convolutional neural networks. Picasso ships with occlusion maps and saliency maps, two visualizations which help reveal issues that evaluation metrics like loss and accuracy might hide: for example, learning a proxy classification task. Picasso works with the Tensorflow deep learning framework, and Keras (when the model can be loaded into the Tensorflow backend. Picasso can be used with minimal configuration by deep learning researchers and engineers alike across various neural network architectures. Adding new visualizations is simple: the user can specify their visualization code and HTML template separately from the application code.

Tutorial on neural network applications in high energy physics: A 1992 perspective

International Nuclear Information System (INIS)

Denby, B.

1992-04-01

Feed forward and recurrent neural networks are introduced and related to standard data analysis tools. Tips are given on applications of neural nets to various areas of high energy physics. A review of applications within high energy physics and a summary of neural net hardware status are given

A stochastic learning algorithm for layered neural networks

International Nuclear Information System (INIS)

Bartlett, E.B.; Uhrig, R.E.

1992-01-01

The random optimization method typically uses a Gaussian probability density function (PDF) to generate a random search vector. In this paper the random search technique is applied to the neural network training problem and is modified to dynamically seek out the optimal probability density function (OPDF) from which to select the search vector. The dynamic OPDF search process, combined with an auto-adaptive stratified sampling technique and a dynamic node architecture (DNA) learning scheme, completes the modifications of the basic method. The DNA technique determines the appropriate number of hidden nodes needed for a given training problem. By using DNA, researchers do not have to set the neural network architectures before training is initiated. The approach is applied to networks of generalized, fully interconnected, continuous perceptions. Computer simulation results are given

Modeling Geomagnetic Variations using a Machine Learning Framework

Science.gov (United States)

Cheung, C. M. M.; Handmer, C.; Kosar, B.; Gerules, G.; Poduval, B.; Mackintosh, G.; Munoz-Jaramillo, A.; Bobra, M.; Hernandez, T.; McGranaghan, R. M.

2017-12-01

We present a framework for data-driven modeling of Heliophysics time series data. The Solar Terrestrial Interaction Neural net Generator (STING) is an open source python module built on top of state-of-the-art statistical learning frameworks (traditional machine learning methods as well as deep learning). To showcase the capability of STING, we deploy it for the problem of predicting the temporal variation of geomagnetic fields. The data used includes solar wind measurements from the OMNI database and geomagnetic field data taken by magnetometers at US Geological Survey observatories. We examine the predictive capability of different machine learning techniques (recurrent neural networks, support vector machines) for a range of forecasting times (minutes to 12 hours). STING is designed to be extensible to other types of data. We show how STING can be used on large sets of data from different sensors/observatories and adapted to tackle other problems in Heliophysics.

Motor sequence learning-induced neural efficiency in functional brain connectivity.

Science.gov (United States)

Karim, Helmet T; Huppert, Theodore J; Erickson, Kirk I; Wollam, Mariegold E; Sparto, Patrick J; Sejdić, Ervin; VanSwearingen, Jessie M

2017-02-15

Previous studies have shown the functional neural circuitry differences before and after an explicitly learned motor sequence task, but have not assessed these changes during the process of motor skill learning. Functional magnetic resonance imaging activity was measured while participants (n=13) were asked to tap their fingers to visually presented sequences in blocks that were either the same sequence repeated (learning block) or random sequences (control block). Motor learning was associated with a decrease in brain activity during learning compared to control. Lower brain activation was noted in the posterior parietal association area and bilateral thalamus during the later periods of learning (not during the control). Compared to the control condition, we found the task-related motor learning was associated with decreased connectivity between the putamen and left inferior frontal gyrus and left middle cingulate brain regions. Motor learning was associated with changes in network activity, spatial extent, and connectivity. Copyright © 2016 Elsevier B.V. All rights reserved.

Distributed Learning, Recognition, and Prediction by ART and ARTMAP Neural Networks.

Science.gov (United States)

Carpenter, Gail A.

1997-11-01

A class of adaptive resonance theory (ART) models for learning, recognition, and prediction with arbitrarily distributed code representations is introduced. Distributed ART neural networks combine the stable fast learning capabilities of winner-take-all ART systems with the noise tolerance and code compression capabilities of multilayer perceptrons. With a winner-take-all code, the unsupervised model dART reduces to fuzzy ART and the supervised model dARTMAP reduces to fuzzy ARTMAP. With a distributed code, these networks automatically apportion learned changes according to the degree of activation of each coding node, which permits fast as well as slow learning without catastrophic forgetting. Distributed ART models replace the traditional neural network path weight with a dynamic weight equal to the rectified difference between coding node activation and an adaptive threshold. Thresholds increase monotonically during learning according to a principle of atrophy due to disuse. However, monotonic change at the synaptic level manifests itself as bidirectional change at the dynamic level, where the result of adaptation resembles long-term potentiation (LTP) for single-pulse or low frequency test inputs but can resemble long-term depression (LTD) for higher frequency test inputs. This paradoxical behavior is traced to dual computational properties of phasic and tonic coding signal components. A parallel distributed match-reset-search process also helps stabilize memory. Without the match-reset-search system, dART becomes a type of distributed competitive learning network.

Neural Correlates of Morphology Acquisition through a Statistical Learning Paradigm.

Science.gov (United States)

Sandoval, Michelle; Patterson, Dianne; Dai, Huanping; Vance, Christopher J; Plante, Elena

2017-01-01

The neural basis of statistical learning as it occurs over time was explored with stimuli drawn from a natural language (Russian nouns). The input reflected the "rules" for marking categories of gendered nouns, without making participants explicitly aware of the nature of what they were to learn. Participants were scanned while listening to a series of gender-marked nouns during four sequential scans, and were tested for their learning immediately after each scan. Although participants were not told the nature of the learning task, they exhibited learning after their initial exposure to the stimuli. Independent component analysis of the brain data revealed five task-related sub-networks. Unlike prior statistical learning studies of word segmentation, this morphological learning task robustly activated the inferior frontal gyrus during the learning period. This region was represented in multiple independent components, suggesting it functions as a network hub for this type of learning. Moreover, the results suggest that subnetworks activated by statistical learning are driven by the nature of the input, rather than reflecting a general statistical learning system.

Universal approximation in p-mean by neural networks

NARCIS (Netherlands)

Burton, R.M; Dehling, H.G

A feedforward neural net with d input neurons and with a single hidden layer of n neurons is given by [GRAPHICS] where a(j), theta(j), w(ji) is an element of R. In this paper we study the approximation of arbitrary functions f: R-d --> R by a neural net in an L-p(mu) norm for some finite measure mu

Deep nets vs expert designed features in medical physics: An IMRT QA case study.

Science.gov (United States)

Interian, Yannet; Rideout, Vincent; Kearney, Vasant P; Gennatas, Efstathios; Morin, Olivier; Cheung, Joey; Solberg, Timothy; Valdes, Gilmer

2018-03-30

The purpose of this study was to compare the performance of Deep Neural Networks against a technique designed by domain experts in the prediction of gamma passing rates for Intensity Modulated Radiation Therapy Quality Assurance (IMRT QA). A total of 498 IMRT plans across all treatment sites were planned in Eclipse version 11 and delivered using a dynamic sliding window technique on Clinac iX or TrueBeam Linacs. Measurements were performed using a commercial 2D diode array, and passing rates for 3%/3 mm local dose/distance-to-agreement (DTA) were recorded. Separately, fluence maps calculated for each plan were used as inputs to a convolution neural network (CNN). The CNNs were trained to predict IMRT QA gamma passing rates using TensorFlow and Keras. A set of model architectures, inspired by the convolutional blocks of the VGG-16 ImageNet model, were constructed and implemented. Synthetic data, created by rotating and translating the fluence maps during training, was created to boost the performance of the CNNs. Dropout, batch normalization, and data augmentation were utilized to help train the model. The performance of the CNNs was compared to a generalized Poisson regression model, previously developed for this application, which used 78 expert designed features. Deep Neural Networks without domain knowledge achieved comparable performance to a baseline system designed by domain experts in the prediction of 3%/3 mm Local gamma passing rates. An ensemble of neural nets resulted in a mean absolute error (MAE) of 0.70 ± 0.05 and the domain expert model resulted in a 0.74 ± 0.06. Convolutional neural networks (CNNs) with transfer learning can predict IMRT QA passing rates by automatically designing features from the fluence maps without human expert supervision. Predictions from CNNs are comparable to a system carefully designed by physicist experts. © 2018 American Association of Physicists in Medicine.

Computational modeling of neural plasticity for self-organization of neural networks.

Science.gov (United States)

Chrol-Cannon, Joseph; Jin, Yaochu

2014-11-01

Self-organization in biological nervous systems during the lifetime is known to largely occur through a process of plasticity that is dependent upon the spike-timing activity in connected neurons. In the field of computational neuroscience, much effort has been dedicated to building up computational models of neural plasticity to replicate experimental data. Most recently, increasing attention has been paid to understanding the role of neural plasticity in functional and structural neural self-organization, as well as its influence on the learning performance of neural networks for accomplishing machine learning tasks such as classification and regression. Although many ideas and hypothesis have been suggested, the relationship between the structure, dynamics and learning performance of neural networks remains elusive. The purpose of this article is to review the most important computational models for neural plasticity and discuss various ideas about neural plasticity's role. Finally, we suggest a few promising research directions, in particular those along the line that combines findings in computational neuroscience and systems biology, and their synergetic roles in understanding learning, memory and cognition, thereby bridging the gap between computational neuroscience, systems biology and computational intelligence. Copyright © 2014 Elsevier Ireland Ltd. All rights reserved.

A bat's ear view of neural nets in physics

International Nuclear Information System (INIS)

Denby, B.

1997-01-01

The use of neural networks in high energy physics has become a field of its own which now has been in existence for ten years. This paper attempts to draw some conclusions on the utility of neural networks for physics applications, and also to make some projections for the future of this line of research. (orig.)

Evaluation of Deep Learning Models for Predicting CO2 Flux

Science.gov (United States)

Halem, M.; Nguyen, P.; Frankel, D.

2017-12-01

Artificial neural networks have been employed to calculate surface flux measurements from station data because they are able to fit highly nonlinear relations between input and output variables without knowing the detail relationships between the variables. However, the accuracy in performing neural net estimates of CO2 flux from observations of CO2 and other atmospheric variables is influenced by the architecture of the neural model, the availability, and complexity of interactions between physical variables such as wind, temperature, and indirect variables like latent heat, and sensible heat, etc. We evaluate two deep learning models, feed forward and recurrent neural network models to learn how they each respond to the physical measurements, time dependency of the measurements of CO2 concentration, humidity, pressure, temperature, wind speed etc. for predicting the CO2 flux. In this paper, we focus on a) building neural network models for estimating CO2 flux based on DOE data from tower Atmospheric Radiation Measurement data; b) evaluating the impact of choosing the surface variables and model hyper-parameters on the accuracy and predictions of surface flux; c) assessing the applicability of the neural network models on estimate CO2 flux by using OCO-2 satellite data; d) studying the efficiency of using GPU-acceleration for neural network performance using IBM Power AI deep learning software and packages on IBM Minsky system.

A Cognitive Neural Architecture Able to Learn and Communicate through Natural Language.

Directory of Open Access Journals (Sweden)

Bruno Golosio

Full Text Available Communicative interactions involve a kind of procedural knowledge that is used by the human brain for processing verbal and nonverbal inputs and for language production. Although considerable work has been done on modeling human language abilities, it has been difficult to bring them together to a comprehensive tabula rasa system compatible with current knowledge of how verbal information is processed in the brain. This work presents a cognitive system, entirely based on a large-scale neural architecture, which was developed to shed light on the procedural knowledge involved in language elaboration. The main component of this system is the central executive, which is a supervising system that coordinates the other components of the working memory. In our model, the central executive is a neural network that takes as input the neural activation states of the short-term memory and yields as output mental actions, which control the flow of information among the working memory components through neural gating mechanisms. The proposed system is capable of learning to communicate through natural language starting from tabula rasa, without any a priori knowledge of the structure of phrases, meaning of words, role of the different classes of words, only by interacting with a human through a text-based interface, using an open-ended incremental learning process. It is able to learn nouns, verbs, adjectives, pronouns and other word classes, and to use them in expressive language. The model was validated on a corpus of 1587 input sentences, based on literature on early language assessment, at the level of about 4-years old child, and produced 521 output sentences, expressing a broad range of language processing functionalities.

Learning to Produce Syllabic Speech Sounds via Reward-Modulated Neural Plasticity

Science.gov (United States)

Warlaumont, Anne S.; Finnegan, Megan K.

2016-01-01

At around 7 months of age, human infants begin to reliably produce well-formed syllables containing both consonants and vowels, a behavior called canonical babbling. Over subsequent months, the frequency of canonical babbling continues to increase. How the infant’s nervous system supports the acquisition of this ability is unknown. Here we present a computational model that combines a spiking neural network, reinforcement-modulated spike-timing-dependent plasticity, and a human-like vocal tract to simulate the acquisition of canonical babbling. Like human infants, the model’s frequency of canonical babbling gradually increases. The model is rewarded when it produces a sound that is more auditorily salient than sounds it has previously produced. This is consistent with data from human infants indicating that contingent adult responses shape infant behavior and with data from deaf and tracheostomized infants indicating that hearing, including hearing one’s own vocalizations, is critical for canonical babbling development. Reward receipt increases the level of dopamine in the neural network. The neural network contains a reservoir with recurrent connections and two motor neuron groups, one agonist and one antagonist, which control the masseter and orbicularis oris muscles, promoting or inhibiting mouth closure. The model learns to increase the number of salient, syllabic sounds it produces by adjusting the base level of muscle activation and increasing their range of activity. Our results support the possibility that through dopamine-modulated spike-timing-dependent plasticity, the motor cortex learns to harness its natural oscillations in activity in order to produce syllabic sounds. It thus suggests that learning to produce rhythmic mouth movements for speech production may be supported by general cortical learning mechanisms. The model makes several testable predictions and has implications for our understanding not only of how syllabic vocalizations develop

Dissociable neural representations of reinforcement and belief prediction errors underlie strategic learning.

Science.gov (United States)

Zhu, Lusha; Mathewson, Kyle E; Hsu, Ming

2012-01-31

Decision-making in the presence of other competitive intelligent agents is fundamental for social and economic behavior. Such decisions require agents to behave strategically, where in addition to learning about the rewards and punishments available in the environment, they also need to anticipate and respond to actions of others competing for the same rewards. However, whereas we know much about strategic learning at both theoretical and behavioral levels, we know relatively little about the underlying neural mechanisms. Here, we show using a multi-strategy competitive learning paradigm that strategic choices can be characterized by extending the reinforcement learning (RL) framework to incorporate agents' beliefs about the actions of their opponents. Furthermore, using this characterization to generate putative internal values, we used model-based functional magnetic resonance imaging to investigate neural computations underlying strategic learning. We found that the distinct notions of prediction errors derived from our computational model are processed in a partially overlapping but distinct set of brain regions. Specifically, we found that the RL prediction error was correlated with activity in the ventral striatum. In contrast, activity in the ventral striatum, as well as the rostral anterior cingulate (rACC), was correlated with a previously uncharacterized belief-based prediction error. Furthermore, activity in rACC reflected individual differences in degree of engagement in belief learning. These results suggest a model of strategic behavior where learning arises from interaction of dissociable reinforcement and belief-based inputs.

Growing adaptive machines combining development and learning in artificial neural networks

CERN Document Server

Bredeche, Nicolas; Doursat, René

2014-01-01

The pursuit of artificial intelligence has been a highly active domain of research for decades, yielding exciting scientific insights and productive new technologies. In terms of generating intelligence, however, this pursuit has yielded only limited success. This book explores the hypothesis that adaptive growth is a means of moving forward. By emulating the biological process of development, we can incorporate desirable characteristics of natural neural systems into engineered designs, and thus move closer towards the creation of brain-like systems. The particular focus is on how to design artificial neural networks for engineering tasks. The book consists of contributions from 18 researchers, ranging from detailed reviews of recent domains by senior scientists, to exciting new contributions representing the state of the art in machine learning research. The book begins with broad overviews of artificial neurogenesis and bio-inspired machine learning, suitable both as an introduction to the domains and as a...

Learning-Related Changes in Adolescents' Neural Networks during Hypothesis-Generating and Hypothesis-Understanding Training

Science.gov (United States)

Lee, Jun-Ki; Kwon, Yongju

2012-01-01

Fourteen science high school students participated in this study, which investigated neural-network plasticity associated with hypothesis-generating and hypothesis-understanding in learning. The students were divided into two groups and participated in either hypothesis-generating or hypothesis-understanding type learning programs, which were…

Supervised Learning in Spiking Neural Networks for Precise Temporal Encoding.

Science.gov (United States)

Gardner, Brian; Grüning, André

2016-01-01

Precise spike timing as a means to encode information in neural networks is biologically supported, and is advantageous over frequency-based codes by processing input features on a much shorter time-scale. For these reasons, much recent attention has been focused on the development of supervised learning rules for spiking neural networks that utilise a temporal coding scheme. However, despite significant progress in this area, there still lack rules that have a theoretical basis, and yet can be considered biologically relevant. Here we examine the general conditions under which synaptic plasticity most effectively takes place to support the supervised learning of a precise temporal code. As part of our analysis we examine two spike-based learning methods: one of which relies on an instantaneous error signal to modify synaptic weights in a network (INST rule), and the other one relying on a filtered error signal for smoother synaptic weight modifications (FILT rule). We test the accuracy of the solutions provided by each rule with respect to their temporal encoding precision, and then measure the maximum number of input patterns they can learn to memorise using the precise timings of individual spikes as an indication of their storage capacity. Our results demonstrate the high performance of the FILT rule in most cases, underpinned by the rule's error-filtering mechanism, which is predicted to provide smooth convergence towards a desired solution during learning. We also find the FILT rule to be most efficient at performing input pattern memorisations, and most noticeably when patterns are identified using spikes with sub-millisecond temporal precision. In comparison with existing work, we determine the performance of the FILT rule to be consistent with that of the highly efficient E-learning Chronotron rule, but with the distinct advantage that our FILT rule is also implementable as an online method for increased biological realism.

Asymmetric Variate Generation via a Parameterless Dual Neural Learning Algorithm

Directory of Open Access Journals (Sweden)

Simone Fiori

2008-01-01

Full Text Available In a previous work (S. Fiori, 2006, we proposed a random number generator based on a tunable non-linear neural system, whose learning rule is designed on the basis of a cardinal equation from statistics and whose implementation is based on look-up tables (LUTs. The aim of the present manuscript is to improve the above-mentioned random number generation method by changing the learning principle, while retaining the efficient LUT-based implementation. The new method proposed here proves easier to implement and relaxes some previous limitations.

Cross-domain and multi-task transfer learning of deep convolutional neural network for breast cancer diagnosis in digital breast tomosynthesis

Science.gov (United States)

Samala, Ravi K.; Chan, Heang-Ping; Hadjiiski, Lubomir; Helvie, Mark A.; Richter, Caleb; Cha, Kenny

2018-02-01

We propose a cross-domain, multi-task transfer learning framework to transfer knowledge learned from non-medical images by a deep convolutional neural network (DCNN) to medical image recognition task while improving the generalization by multi-task learning of auxiliary tasks. A first stage cross-domain transfer learning was initiated from ImageNet trained DCNN to mammography trained DCNN. 19,632 regions-of-interest (ROI) from 2,454 mass lesions were collected from two imaging modalities: digitized-screen film mammography (SFM) and full-field digital mammography (DM), and split into training and test sets. In the multi-task transfer learning, the DCNN learned the mass classification task simultaneously from the training set of SFM and DM. The best transfer network for mammography was selected from three transfer networks with different number of convolutional layers frozen. The performance of single-task and multitask transfer learning on an independent SFM test set in terms of the area under the receiver operating characteristic curve (AUC) was 0.78+/-0.02 and 0.82+/-0.02, respectively. In the second stage cross-domain transfer learning, a set of 12,680 ROIs from 317 mass lesions on DBT were split into validation and independent test sets. We first studied the data requirements for the first stage mammography trained DCNN by varying the mammography training data from 1% to 100% and evaluated its learning on the DBT validation set in inference mode. We found that the entire available mammography set provided the best generalization. The DBT validation set was then used to train only the last four fully connected layers, resulting in an AUC of 0.90+/-0.04 on the independent DBT test set.

Optical implementation of neural learning algorithms based on cross-gain modulation in a semiconductor optical amplifier

Science.gov (United States)

Li, Qiang; Wang, Zhi; Le, Yansi; Sun, Chonghui; Song, Xiaojia; Wu, Chongqing

2016-10-01

Neuromorphic engineering has a wide range of applications in the fields of machine learning, pattern recognition, adaptive control, etc. Photonics, characterized by its high speed, wide bandwidth, low power consumption and massive parallelism, is an ideal way to realize ultrafast spiking neural networks (SNNs). Synaptic plasticity is believed to be critical for learning, memory and development in neural circuits. Experimental results have shown that changes of synapse are highly dependent on the relative timing of pre- and postsynaptic spikes. Synaptic plasticity in which presynaptic spikes preceding postsynaptic spikes results in strengthening, while the opposite timing results in weakening is called antisymmetric spike-timing-dependent plasticity (STDP) learning rule. And synaptic plasticity has the opposite effect under the same conditions is called antisymmetric anti-STDP learning rule. We proposed and experimentally demonstrated an optical implementation of neural learning algorithms, which can achieve both of antisymmetric STDP and anti-STDP learning rule, based on the cross-gain modulation (XGM) within a single semiconductor optical amplifier (SOA). The weight and height of the potentitation and depression window can be controlled by adjusting the injection current of the SOA, to mimic the biological antisymmetric STDP and anti-STDP learning rule more realistically. As the injection current increases, the width of depression and potentitation window decreases and height increases, due to the decreasing of recovery time and increasing of gain under a stronger injection current. Based on the demonstrated optical STDP circuit, ultrafast learning in optical SNNs can be realized.

Color image definition evaluation method based on deep learning method

Science.gov (United States)

Liu, Di; Li, YingChun

2018-01-01

In order to evaluate different blurring levels of color image and improve the method of image definition evaluation, this paper proposed a method based on the depth learning framework and BP neural network classification model, and presents a non-reference color image clarity evaluation method. Firstly, using VGG16 net as the feature extractor to extract 4,096 dimensions features of the images, then the extracted features and labeled images are employed in BP neural network to train. And finally achieve the color image definition evaluation. The method in this paper are experimented by using images from the CSIQ database. The images are blurred at different levels. There are 4,000 images after the processing. Dividing the 4,000 images into three categories, each category represents a blur level. 300 out of 400 high-dimensional features are trained in VGG16 net and BP neural network, and the rest of 100 samples are tested. The experimental results show that the method can take full advantage of the learning and characterization capability of deep learning. Referring to the current shortcomings of the major existing image clarity evaluation methods, which manually design and extract features. The method in this paper can extract the images features automatically, and has got excellent image quality classification accuracy for the test data set. The accuracy rate is 96%. Moreover, the predicted quality levels of original color images are similar to the perception of the human visual system.

Ventral Tegmental Area and Substantia Nigra Neural Correlates of Spatial Learning

Science.gov (United States)

Martig, Adria K.; Mizumori, Sheri J. Y.

2011-01-01

The ventral tegmental area (VTA) and substantia nigra pars compacta (SNc) may provide modulatory signals that, respectively, influence hippocampal (HPC)- and striatal-dependent memory. Electrophysiological studies investigating neural correlates of learning and memory of dopamine (DA) neurons during classical conditioning tasks have found DA…

The neural coding of feedback learning across child and adolescent development

NARCIS (Netherlands)

Peters, S.; Braams, B.R.; Raijmakers, M.E.J.; Koolschijn, P.C.M.P.; Crone, E.A.

2014-01-01

The ability to learn from environmental cues is an important contributor to successful performance in a variety of settings, including school. Despite the progress in unraveling the neural correlates of cognitive control in childhood and adolescence, relatively little is known about how these brain

Neural changes associated to procedural learning and automatization process in Developmental Coordination Disorder and/or Developmental Dyslexia.

Science.gov (United States)

Biotteau, Maëlle; Péran, Patrice; Vayssière, Nathalie; Tallet, Jessica; Albaret, Jean-Michel; Chaix, Yves

2017-03-01

Recent theories hypothesize that procedural learning may support the frequent overlap between neurodevelopmental disorders. The neural circuitry supporting procedural learning includes, among others, cortico-cerebellar and cortico-striatal loops. Alteration of these loops may account for the frequent comorbidity between Developmental Coordination Disorder (DCD) and Developmental Dyslexia (DD). The aim of our study was to investigate cerebral changes due to the learning and automatization of a sequence learning task in children with DD, or DCD, or both disorders. fMRI on 48 children (aged 8-12) with DD, DCD or DD + DCD was used to explore their brain activity during procedural tasks, performed either after two weeks of training or in the early stage of learning. Firstly, our results indicate that all children were able to perform the task with the same level of automaticity, but recruit different brain processes to achieve the same performance. Secondly, our fMRI results do not appear to confirm Nicolson and Fawcett's model. The neural correlates recruited for procedural learning by the DD and the comorbid groups are very close, while the DCD group presents distinct characteristics. This provide a promising direction on the neural mechanisms associated with procedural learning in neurodevelopmental disorders and for understanding comorbidity. Published by Elsevier Ltd.

Neural Network Machine Learning and Dimension Reduction for Data Visualization

Science.gov (United States)

Liles, Charles A.

2014-01-01

Neural network machine learning in computer science is a continuously developing field of study. Although neural network models have been developed which can accurately predict a numeric value or nominal classification, a general purpose method for constructing neural network architecture has yet to be developed. Computer scientists are often forced to rely on a trial-and-error process of developing and improving accurate neural network models. In many cases, models are constructed from a large number of input parameters. Understanding which input parameters have the greatest impact on the prediction of the model is often difficult to surmise, especially when the number of input variables is very high. This challenge is often labeled the "curse of dimensionality" in scientific fields. However, techniques exist for reducing the dimensionality of problems to just two dimensions. Once a problem's dimensions have been mapped to two dimensions, it can be easily plotted and understood by humans. The ability to visualize a multi-dimensional dataset can provide a means of identifying which input variables have the highest effect on determining a nominal or numeric output. Identifying these variables can provide a better means of training neural network models; models can be more easily and quickly trained using only input variables which appear to affect the outcome variable. The purpose of this project is to explore varying means of training neural networks and to utilize dimensional reduction for visualizing and understanding complex datasets.

Adaptive Near-Optimal Multiuser Detection Using a Stochastic and Hysteretic Hopfield Net Receiver

Directory of Open Access Journals (Sweden)

Gábor Jeney

2003-01-01

Full Text Available This paper proposes a novel adaptive MUD algorithm for a wide variety (practically any kind of interference limited systems, for example, code division multiple access (CDMA. The algorithm is based on recently developed neural network techniques and can perform near optimal detection in the case of unknown channel characteristics. The proposed algorithm consists of two main blocks; one estimates the symbols sent by the transmitters, the other identifies each channel of the corresponding communication links. The estimation of symbols is carried out either by a stochastic Hopfield net (SHN or by a hysteretic neural network (HyNN or both. The channel identification is based on either the self-organizing feature map (SOM or the learning vector quantization (LVQ. The combination of these two blocks yields a powerful real-time detector with near optimal performance. The performance is analyzed by extensive simulations.

Dynamic Learning from Adaptive Neural Control of Uncertain Robots with Guaranteed Full-State Tracking Precision

Directory of Open Access Journals (Sweden)

Min Wang

2017-01-01

Full Text Available A dynamic learning method is developed for an uncertain n-link robot with unknown system dynamics, achieving predefined performance attributes on the link angular position and velocity tracking errors. For a known nonsingular initial robotic condition, performance functions and unconstrained transformation errors are employed to prevent the violation of the full-state tracking error constraints. By combining two independent Lyapunov functions and radial basis function (RBF neural network (NN approximator, a novel and simple adaptive neural control scheme is proposed for the dynamics of the unconstrained transformation errors, which guarantees uniformly ultimate boundedness of all the signals in the closed-loop system. In the steady-state control process, RBF NNs are verified to satisfy the partial persistent excitation (PE condition. Subsequently, an appropriate state transformation is adopted to achieve the accurate convergence of neural weight estimates. The corresponding experienced knowledge on unknown robotic dynamics is stored in NNs with constant neural weight values. Using the stored knowledge, a static neural learning controller is developed to improve the full-state tracking performance. A comparative simulation study on a 2-link robot illustrates the effectiveness of the proposed scheme.

Deep-Learning Convolutional Neural Networks Accurately Classify Genetic Mutations in Gliomas.

Science.gov (United States)

Chang, P; Grinband, J; Weinberg, B D; Bardis, M; Khy, M; Cadena, G; Su, M-Y; Cha, S; Filippi, C G; Bota, D; Baldi, P; Poisson, L M; Jain, R; Chow, D

2018-05-10

The World Health Organization has recently placed new emphasis on the integration of genetic information for gliomas. While tissue sampling remains the criterion standard, noninvasive imaging techniques may provide complimentary insight into clinically relevant genetic mutations. Our aim was to train a convolutional neural network to independently predict underlying molecular genetic mutation status in gliomas with high accuracy and identify the most predictive imaging features for each mutation. MR imaging data and molecular information were retrospectively obtained from The Cancer Imaging Archives for 259 patients with either low- or high-grade gliomas. A convolutional neural network was trained to classify isocitrate dehydrogenase 1 ( IDH1 ) mutation status, 1p/19q codeletion, and O6-methylguanine-DNA methyltransferase ( MGMT ) promotor methylation status. Principal component analysis of the final convolutional neural network layer was used to extract the key imaging features critical for successful classification. Classification had high accuracy: IDH1 mutation status, 94%; 1p/19q codeletion, 92%; and MGMT promotor methylation status, 83%. Each genetic category was also associated with distinctive imaging features such as definition of tumor margins, T1 and FLAIR suppression, extent of edema, extent of necrosis, and textural features. Our results indicate that for The Cancer Imaging Archives dataset, machine-learning approaches allow classification of individual genetic mutations of both low- and high-grade gliomas. We show that relevant MR imaging features acquired from an added dimensionality-reduction technique demonstrate that neural networks are capable of learning key imaging components without prior feature selection or human-directed training. © 2018 by American Journal of Neuroradiology.

Biologically plausible learning in neural networks: a lesson from bacterial chemotaxis.

Science.gov (United States)

Shimansky, Yury P

2009-12-01

Learning processes in the brain are usually associated with plastic changes made to optimize the strength of connections between neurons. Although many details related to biophysical mechanisms of synaptic plasticity have been discovered, it is unclear how the concurrent performance of adaptive modifications in a huge number of spatial locations is organized to minimize a given objective function. Since direct experimental observation of even a relatively small subset of such changes is not feasible, computational modeling is an indispensable investigation tool for solving this problem. However, the conventional method of error back-propagation (EBP) employed for optimizing synaptic weights in artificial neural networks is not biologically plausible. This study based on computational experiments demonstrated that such optimization can be performed rather efficiently using the same general method that bacteria employ for moving closer to an attractant or away from a repellent. With regard to neural network optimization, this method consists of regulating the probability of an abrupt change in the direction of synaptic weight modification according to the temporal gradient of the objective function. Neural networks utilizing this method (regulation of modification probability, RMP) can be viewed as analogous to swimming in the multidimensional space of their parameters in the flow of biochemical agents carrying information about the optimality criterion. The efficiency of RMP is comparable to that of EBP, while RMP has several important advantages. Since the biological plausibility of RMP is beyond a reasonable doubt, the RMP concept provides a constructive framework for the experimental analysis of learning in natural neural networks.

Automated sleep stage detection with a classical and a neural learning algorithm--methodological aspects.

Science.gov (United States)

Schwaibold, M; Schöchlin, J; Bolz, A

2002-01-01

For classification tasks in biosignal processing, several strategies and algorithms can be used. Knowledge-based systems allow prior knowledge about the decision process to be integrated, both by the developer and by self-learning capabilities. For the classification stages in a sleep stage detection framework, three inference strategies were compared regarding their specific strengths: a classical signal processing approach, artificial neural networks and neuro-fuzzy systems. Methodological aspects were assessed to attain optimum performance and maximum transparency for the user. Due to their effective and robust learning behavior, artificial neural networks could be recommended for pattern recognition, while neuro-fuzzy systems performed best for the processing of contextual information.

The Use of Neural Network to Recognize the Parts of the Computer Motherboard

OpenAIRE

Abbas M. Ali; S. D. Gore; Musaab AL-Sarierah

2005-01-01

This study suggests a new approach of learning which utilizes the techniques of computer vision to recognize the parts inside the motherboard. The main thrust is to identify different parts of the motherboard using a Hopfield Neural Network. The outcome of the net is compared with the objects stored in the database. The proposed scheme is implemented using bottom -up approach, where steps like edge detection, spatial filtering, image masking..etc are performed in sequence. the scheme is simul...

The interchangeability of learning rate and gain in backpropagation neural networks

NARCIS (Netherlands)

Thimm, G.; Moerland, P.; Fiesler, E.

1996-01-01

The backpropagation algorithm is widely used for training multilayer neural networks. In this publication the gain of its activation function(s) is investigated. In specific, it is proven that changing the gain of the activation function is equivalent to changing the learning rate and the weights.

Development of a neural net paradigm that predicts simulator sickness

Energy Technology Data Exchange (ETDEWEB)

Allgood, G.O.

1993-03-01

A disease exists that affects pilots and aircrew members who use Navy Operational Flight Training Systems. This malady, commonly referred to as simulator sickness and whose symptomatology closely aligns with that of motion sickness, can compromise the use of these systems because of a reduced utilization factor, negative transfer of training, and reduction in combat readiness. A report is submitted that develops an artificial neural network (ANN) and behavioral model that predicts the onset and level of simulator sickness in the pilots and aircrews who sue these systems. It is proposed that the paradigm could be implemented in real time as a biofeedback monitor to reduce the risk to users of these systems. The model captures the neurophysiological impact of use (human-machine interaction) by developing a structure that maps the associative and nonassociative behavioral patterns (learned expectations) and vestibular (otolith and semicircular canals of the inner ear) and tactile interaction, derived from system acceleration profiles, onto an abstract space that predicts simulator sickness for a given training flight.

Turkey's net energy consumption

International Nuclear Information System (INIS)

Soezen, Adnan; Arcaklioglu, Erol; Oezkaymak, Mehmet

2005-01-01

The main goal of this study is to develop the equations for forecasting net energy consumption (NEC) using an artificial neural-network (ANN) technique in order to determine the future level of energy consumption in Turkey. In this study, two different models were used in order to train the neural network. In one of them, population, gross generation, installed capacity and years are used in the input layer of the network (Model 1). Other energy sources are used in input layer of network (Model 2). The net energy consumption is in the output layer for two models. Data from 1975 to 2003 are used for the training. Three years (1981, 1994 and 2003) are used only as test data to confirm this method. The statistical coefficients of multiple determinations (R 2 -value) for training data are equal to 0.99944 and 0.99913 for Models 1 and 2, respectively. Similarly, R 2 values for testing data are equal to 0.997386 and 0.999558 for Models 1 and 2, respectively. According to the results, the net energy consumption using the ANN technique has been predicted with acceptable accuracy. Apart from reducing the whole time required, with the ANN approach, it is possible to find solutions that make energy applications more viable and thus more attractive to potential users. It is also expected that this study will be helpful in developing highly applicable energy policies

Neural Network for Image-to-Image Control of Optical Tweezers

Science.gov (United States)

Decker, Arthur J.; Anderson, Robert C.; Weiland, Kenneth E.; Wrbanek, Susan Y.

2004-01-01

A method is discussed for using neural networks to control optical tweezers. Neural-net outputs are combined with scaling and tiling to generate 480 by 480-pixel control patterns for a spatial light modulator (SLM). The SLM can be combined in various ways with a microscope to create movable tweezers traps with controllable profiles. The neural nets are intended to respond to scattered light from carbon and silicon carbide nanotube sensors. The nanotube sensors are to be held by the traps for manipulation and calibration. Scaling and tiling allow the 100 by 100-pixel maximum resolution of the neural-net software to be applied in stages to exploit the full 480 by 480-pixel resolution of the SLM. One of these stages is intended to create sensitive null detectors for detecting variations in the scattered light from the nanotube sensors.

Optimal Search Strategy of Robotic Assembly Based on Neural Vibration Learning

Directory of Open Access Journals (Sweden)

Lejla Banjanovic-Mehmedovic

2011-01-01

Full Text Available This paper presents implementation of optimal search strategy (OSS in verification of assembly process based on neural vibration learning. The application problem is the complex robot assembly of miniature parts in the example of mating the gears of one multistage planetary speed reducer. Assembly of tube over the planetary gears was noticed as the most difficult problem of overall assembly. The favourable influence of vibration and rotation movement on compensation of tolerance was also observed. With the proposed neural-network-based learning algorithm, it is possible to find extended scope of vibration state parameter. Using optimal search strategy based on minimal distance path between vibration parameter stage sets (amplitude and frequencies of robots gripe vibration and recovery parameter algorithm, we can improve the robot assembly behaviour, that is, allow the fastest possible way of mating. We have verified by using simulation programs that search strategy is suitable for the situation of unexpected events due to uncertainties.

MosquitoNet: investigating the use of UAV and artificial neural networks for integrated mosquito management

Science.gov (United States)

Case, E.; Ren, Y.; Shragai, T.; Erickson, D.

2017-12-01

Integrated mosquito control is expensive and resource intensive, and changing climatic factors are predicted to expand the habitable range of disease-carrying mosquitoes into new regions in the United States. Of particular concern in the northeastern United States are aedes albopictus, an aggressive, invasive species of mosquito that can transmit both native and exotic disease. Ae. albopictus prefer to live near human populations and breed in artificial containers with as little as two millimeters of standing water, exponentially increasing the difficulty of source control in suburban and urban areas. However, low-cost unmanned aerial vehicles (UAVs) can be used to photograph large regions at centimeter-resolution, and can image containers of interest in suburban neighborhoods. While proofs-of-concepts have been shown using UAVs to identify naturally occurring bodies of water, they have not been used to identify mosquito habitat in more populated areas. One of the primary challenges is that post-processing high-resolution aerial imagery is still time intensive, often labelled by hand or with programs built for satellite imagery. Artificial neural networks have been highly successful at image recognition tasks; in the past five years, convolutional neural networks (CNN) have surpassed or aided trained humans in identification of skin cancer, agricultural crops, and poverty levels from satellite imagery. MosquitoNet, a dual classifier built from the Single Shot Multibox Detector and VGG16 architectures, was trained on UAV­­­­­ aerial imagery taken during a larval study in Westchester County in southern New York State in July and August 2017. MosquitoNet was designed to assess the habitat risk of suburban properties by automating the identification and counting of containers like tires, toys, garbage bins, flower pots, etc. The SSD-based architecture marked small containers and other habitat indicators while the VGG16-based architecture classified the type of

Forecasting financial asset processes: stochastic dynamics via learning neural networks.

Science.gov (United States)

Giebel, S; Rainer, M

2010-01-01

Models for financial asset dynamics usually take into account their inherent unpredictable nature by including a suitable stochastic component into their process. Unknown (forward) values of financial assets (at a given time in the future) are usually estimated as expectations of the stochastic asset under a suitable risk-neutral measure. This estimation requires the stochastic model to be calibrated to some history of sufficient length in the past. Apart from inherent limitations, due to the stochastic nature of the process, the predictive power is also limited by the simplifying assumptions of the common calibration methods, such as maximum likelihood estimation and regression methods, performed often without weights on the historic time series, or with static weights only. Here we propose a novel method of "intelligent" calibration, using learning neural networks in order to dynamically adapt the parameters of the stochastic model. Hence we have a stochastic process with time dependent parameters, the dynamics of the parameters being themselves learned continuously by a neural network. The back propagation in training the previous weights is limited to a certain memory length (in the examples we consider 10 previous business days), which is similar to the maximal time lag of autoregressive processes. We demonstrate the learning efficiency of the new algorithm by tracking the next-day forecasts for the EURTRY and EUR-HUF exchange rates each.

A role for adult TLX-positive neural stem cells in learning and behaviour.

Science.gov (United States)

Zhang, Chun-Li; Zou, Yuhua; He, Weimin; Gage, Fred H; Evans, Ronald M

2008-02-21

Neurogenesis persists in the adult brain and can be regulated by a plethora of external stimuli, such as learning, memory, exercise, environment and stress. Although newly generated neurons are able to migrate and preferentially incorporate into the neural network, how these cells are molecularly regulated and whether they are required for any normal brain function are unresolved questions. The adult neural stem cell pool is composed of orphan nuclear receptor TLX-positive cells. Here, using genetic approaches in mice, we demonstrate that TLX (also called NR2E1) regulates adult neural stem cell proliferation in a cell-autonomous manner by controlling a defined genetic network implicated in cell proliferation and growth. Consequently, specific removal of TLX from the adult mouse brain through inducible recombination results in a significant reduction of stem cell proliferation and a marked decrement in spatial learning. In contrast, the resulting suppression of adult neurogenesis does not affect contextual fear conditioning, locomotion or diurnal rhythmic activities, indicating a more selective contribution of newly generated neurons to specific cognitive functions.

Stochastic sensitivity analysis and Langevin simulation for neural network learning

International Nuclear Information System (INIS)

Koda, Masato

1997-01-01

A comprehensive theoretical framework is proposed for the learning of a class of gradient-type neural networks with an additive Gaussian white noise process. The study is based on stochastic sensitivity analysis techniques, and formal expressions are obtained for stochastic learning laws in terms of functional derivative sensitivity coefficients. The present method, based on Langevin simulation techniques, uses only the internal states of the network and ubiquitous noise to compute the learning information inherent in the stochastic correlation between noise signals and the performance functional. In particular, the method does not require the solution of adjoint equations of the back-propagation type. Thus, the present algorithm has the potential for efficiently learning network weights with significantly fewer computations. Application to an unfolded multi-layered network is described, and the results are compared with those obtained by using a back-propagation method

Adaptive neural network/expert system that learns fault diagnosis for different structures

Science.gov (United States)

Simon, Solomon H.

1992-08-01

Corporations need better real-time monitoring and control systems to improve productivity by watching quality and increasing production flexibility. The innovative technology to achieve this goal is evolving in the form artificial intelligence and neural networks applied to sensor processing, fusion, and interpretation. By using these advanced Al techniques, we can leverage existing systems and add value to conventional techniques. Neural networks and knowledge-based expert systems can be combined into intelligent sensor systems which provide real-time monitoring, control, evaluation, and fault diagnosis for production systems. Neural network-based intelligent sensor systems are more reliable because they can provide continuous, non-destructive monitoring and inspection. Use of neural networks can result in sensor fusion and the ability to model highly, non-linear systems. Improved models can provide a foundation for more accurate performance parameters and predictions. We discuss a research software/hardware prototype which integrates neural networks, expert systems, and sensor technologies and which can adapt across a variety of structures to perform fault diagnosis. The flexibility and adaptability of the prototype in learning two structures is presented. Potential applications are discussed.

SEMANTIC3D.NET: a New Large-Scale Point Cloud Classification Benchmark

Science.gov (United States)

Hackel, T.; Savinov, N.; Ladicky, L.; Wegner, J. D.; Schindler, K.; Pollefeys, M.

2017-05-01

This paper presents a new 3D point cloud classification benchmark data set with over four billion manually labelled points, meant as input for data-hungry (deep) learning methods. We also discuss first submissions to the benchmark that use deep convolutional neural networks (CNNs) as a work horse, which already show remarkable performance improvements over state-of-the-art. CNNs have become the de-facto standard for many tasks in computer vision and machine learning like semantic segmentation or object detection in images, but have no yet led to a true breakthrough for 3D point cloud labelling tasks due to lack of training data. With the massive data set presented in this paper, we aim at closing this data gap to help unleash the full potential of deep learning methods for 3D labelling tasks. Our semantic3D.net data set consists of dense point clouds acquired with static terrestrial laser scanners. It contains 8 semantic classes and covers a wide range of urban outdoor scenes: churches, streets, railroad tracks, squares, villages, soccer fields and castles. We describe our labelling interface and show that our data set provides more dense and complete point clouds with much higher overall number of labelled points compared to those already available to the research community. We further provide baseline method descriptions and comparison between methods submitted to our online system. We hope semantic3D.net will pave the way for deep learning methods in 3D point cloud labelling to learn richer, more general 3D representations, and first submissions after only a few months indicate that this might indeed be the case.

Incremental learning of perceptual and conceptual representations and the puzzle of neural repetition suppression.

Science.gov (United States)

Gotts, Stephen J

2016-08-01

Incremental learning models of long-term perceptual and conceptual knowledge hold that neural representations are gradually acquired over many individual experiences via Hebbian-like activity-dependent synaptic plasticity across cortical connections of the brain. In such models, variation in task relevance of information, anatomic constraints, and the statistics of sensory inputs and motor outputs lead to qualitative alterations in the nature of representations that are acquired. Here, the proposal that behavioral repetition priming and neural repetition suppression effects are empirical markers of incremental learning in the cortex is discussed, and research results that both support and challenge this position are reviewed. Discussion is focused on a recent fMRI-adaptation study from our laboratory that shows decoupling of experience-dependent changes in neural tuning, priming, and repetition suppression, with representational changes that appear to work counter to the explicit task demands. Finally, critical experiments that may help to clarify and resolve current challenges are outlined.

Comparisons of a Quantum Annealing and Classical Computer Neural Net Approach for Inferring Global Annual CO2 Fluxes over Land

Science.gov (United States)

Halem, M.; Radov, A.; Singh, D.

2017-12-01

Investigations of mid to high latitude atmospheric CO2 show growing amplitudes in seasonal variations over the past several decades. Recent high-resolution satellite measurements of CO2 concentration are now available for three years from the Orbiting Carbon Observatory-2. The Atmospheric Radiation Measurement (ARM) program of DOE has been making long-term CO2-flux measurements (in addition to CO2 concentration and an array of other meteorological quantities) at several towers and mobile sites located around the globe at half-hour frequencies. Recent papers have shown CO2 fluxes inferred by assimilating CO2 observations into ecosystem models are largely inconsistent with station observations. An investigation of how the biosphere has reacted to changes in atmospheric CO2 is essential to our understanding of potential climate-vegetation feedbacks. Thus, new approaches for calculating CO2-flux for assimilation into land surface models are necessary for improving the prediction of annual carbon uptake. In this study, we calculate and compare the predicted CO2 fluxes results employing a Feed Forward Backward Propagation Neural Network model on two architectures, (i) an IBM Minsky Computer node and (ii) a hybrid version of the ARC D-Wave quantum annealing computer. We compare the neural net results of predictions of CO2 flux from ARM station data for three different DOE ecosystem sites; an arid plains near Oklahoma City, a northern arctic site at Barrows AL, and a tropical rainforest site in the Amazon. Training times and predictive results for the calculating annual CO2 flux for the two architectures for each of the three sites are presented. Comparative results of predictions as measured by RMSE and MAE are discussed. Plots and correlations of observed vs predicted CO2 flux are also presented for all three sites. We show the estimated training times for quantum and classical calculations when extended to calculating global annual Carbon Uptake over land. We also

Scaling deep learning workloads: NVIDIA DGX-1/Pascal and Intel Knights Landing

Energy Technology Data Exchange (ETDEWEB)

Gawande, Nitin A.; Landwehr, Joshua B.; Daily, Jeffrey A.; Tallent, Nathan R.; Vishnu, Abhinav; Kerbyson, Darren J.

2017-08-24

Deep Learning (DL) algorithms have become ubiquitous in data analytics. As a result, major computing vendors --- including NVIDIA, Intel, AMD, and IBM --- have architectural road-maps influenced by DL workloads. Furthermore, several vendors have recently advertised new computing products as accelerating large DL workloads. Unfortunately, it is difficult for data scientists to quantify the potential of these different products. This paper provides a performance and power analysis of important DL workloads on two major parallel architectures: NVIDIA DGX-1 (eight Pascal P100 GPUs interconnected with NVLink) and Intel Knights Landing (KNL) CPUs interconnected with Intel Omni-Path or Cray Aries. Our evaluation consists of a cross section of convolutional neural net workloads: CifarNet, AlexNet, GoogLeNet, and ResNet50 topologies using the Cifar10 and ImageNet datasets. The workloads are vendor-optimized for each architecture. Our analysis indicates that although GPUs provide the highest overall performance, the gap can close for some convolutional networks; and the KNL can be competitive in performance/watt. We find that NVLink facilitates scaling efficiency on GPUs. However, its importance is heavily dependent on neural network architecture. Furthermore, for weak-scaling --- sometimes encouraged by restricted GPU memory --- NVLink is less important.

Residual Deep Convolutional Neural Network Predicts MGMT Methylation Status.

Science.gov (United States)

Korfiatis, Panagiotis; Kline, Timothy L; Lachance, Daniel H; Parney, Ian F; Buckner, Jan C; Erickson, Bradley J

2017-10-01

Predicting methylation of the O6-methylguanine methyltransferase (MGMT) gene status utilizing MRI imaging is of high importance since it is a predictor of response and prognosis in brain tumors. In this study, we compare three different residual deep neural network (ResNet) architectures to evaluate their ability in predicting MGMT methylation status without the need for a distinct tumor segmentation step. We found that the ResNet50 (50 layers) architecture was the best performing model, achieving an accuracy of 94.90% (+/- 3.92%) for the test set (classification of a slice as no tumor, methylated MGMT, or non-methylated). ResNet34 (34 layers) achieved 80.72% (+/- 13.61%) while ResNet18 (18 layers) accuracy was 76.75% (+/- 20.67%). ResNet50 performance was statistically significantly better than both ResNet18 and ResNet34 architectures (p deep neural architectures can be used to predict molecular biomarkers from routine medical images.

Lessons Learned from Net Zero Energy Assessments and Renewable Energy Projects at Military Installations

Energy Technology Data Exchange (ETDEWEB)

Callahan, M.; Anderson, K.; Booth, S.; Katz, J.; Tetreault, T.

2011-09-01

Report highlights the increase in resources, project speed, and scale that is required to achieve the U.S. Department of Defense (DoD) energy efficiency and renewable energy goals and summarizes the net zero energy installation assessment (NZEI) process and the lessons learned from NZEI assessments and large-scale renewable energy projects implementations at DoD installations.

Application of Convolution Neural Network to the forecasts of flare classification and occurrence using SOHO MDI data

Science.gov (United States)

Park, Eunsu; Moon, Yong-Jae

2017-08-01

A Convolutional Neural Network(CNN) is one of the well-known deep-learning methods in image processing and computer vision area. In this study, we apply CNN to two kinds of flare forecasting models: flare classification and occurrence. For this, we consider several pre-trained models (e.g., AlexNet, GoogLeNet, and ResNet) and customize them by changing several options such as the number of layers, activation function, and optimizer. Our inputs are the same number of SOHO)/MDI images for each flare class (None, C, M and X) at 00:00 UT from Jan 1996 to Dec 2010 (total 1600 images). Outputs are the results of daily flare forecasting for flare class and occurrence. We build, train, and test the models on TensorFlow, which is well-known machine learning software library developed by Google. Our major results from this study are as follows. First, most of the models have accuracies more than 0.7. Second, ResNet developed by Microsoft has the best accuracies : 0.86 for flare classification and 0.84 for flare occurrence. Third, the accuracies of these models vary greatly with changing parameters. We discuss several possibilities to improve the models.

Application of neural networks to seismic active control

International Nuclear Information System (INIS)

Tang, Yu.

1995-01-01

An exploratory study on seismic active control using an artificial neural network (ANN) is presented in which a singledegree-of-freedom (SDF) structural system is controlled by a trained neural network. A feed-forward neural network and the backpropagation training method are used in the study. In backpropagation training, the learning rate is determined by ensuring the decrease of the error function at each training cycle. The training patterns for the neural net are generated randomly. Then, the trained ANN is used to compute the control force according to the control algorithm. The control strategy proposed herein is to apply the control force at every time step to destroy the build-up of the system response. The ground motions considered in the simulations are the N21E and N69W components of the Lake Hughes No. 12 record that occurred in the San Fernando Valley in California on February 9, 1971. Significant reduction of the structural response by one order of magnitude is observed. Also, it is shown that the proposed control strategy has the ability to reduce the peak that occurs during the first few cycles of the time history. These promising results assert the potential of applying ANNs to active structural control under seismic loads

Chaos Synchronization Using Adaptive Dynamic Neural Network Controller with Variable Learning Rates

Directory of Open Access Journals (Sweden)

Chih-Hong Kao

2011-01-01

Full Text Available This paper addresses the synchronization of chaotic gyros with unknown parameters and external disturbance via an adaptive dynamic neural network control (ADNNC system. The proposed ADNNC system is composed of a neural controller and a smooth compensator. The neural controller uses a dynamic RBF (DRBF network to online approximate an ideal controller. The DRBF network can create new hidden neurons online if the input data falls outside the hidden layer and prune the insignificant hidden neurons online if the hidden neuron is inappropriate. The smooth compensator is designed to compensate for the approximation error between the neural controller and the ideal controller. Moreover, the variable learning rates of the parameter adaptation laws are derived based on a discrete-type Lyapunov function to speed up the convergence rate of the tracking error. Finally, the simulation results which verified the chaotic behavior of two nonlinear identical chaotic gyros can be synchronized using the proposed ADNNC scheme.

Sentiment analysis: a comparison of deep learning neural network algorithm with SVM and naϊve Bayes for Indonesian text

Science.gov (United States)

Calvin Frans Mariel, Wahyu; Mariyah, Siti; Pramana, Setia

2018-03-01

Deep learning is a new era of machine learning techniques that essentially imitate the structure and function of the human brain. It is a development of deeper Artificial Neural Network (ANN) that uses more than one hidden layer. Deep Learning Neural Network has a great ability on recognizing patterns from various data types such as picture, audio, text, and many more. In this paper, the authors tries to measure that algorithm’s ability by applying it into the text classification. The classification task herein is done by considering the content of sentiment in a text which is also called as sentiment analysis. By using several combinations of text preprocessing and feature extraction techniques, we aim to compare the precise modelling results of Deep Learning Neural Network with the other two commonly used algorithms, the Naϊve Bayes and Support Vector Machine (SVM). This algorithm comparison uses Indonesian text data with balanced and unbalanced sentiment composition. Based on the experimental simulation, Deep Learning Neural Network clearly outperforms the Naϊve Bayes and SVM and offers a better F-1 Score while for the best feature extraction technique which improves that modelling result is Bigram.

Have we met before? Neural correlates of emotional learning in women with social phobia.

Science.gov (United States)

Laeger, Inga; Keuper, Kati; Heitmann, Carina; Kugel, Harald; Dobel, Christian; Eden, Annuschka; Arolt, Volker; Zwitserlood, Pienie; Dannlowski, Udo; Zwanzger, Peter

2014-05-01

Altered memory processes are thought to be a key mechanism in the etiology of anxiety disorders, but little is known about the neural correlates of fear learning and memory biases in patients with social phobia. The present study therefore examined whether patients with social phobia exhibit different patterns of neural activation when confronted with recently acquired emotional stimuli. Patients with social phobia and a group of healthy controls learned to associate pseudonames with pictures of persons displaying either a fearful or a neutral expression. The next day, participants read the pseudonames in the magnetic resonance imaging scanner. Afterwards, 2 memory tests were carried out. We enrolled 21 patients and 21 controls in our study. There were no group differences for learning performance, and results of the memory tests were mixed. On a neural level, patients showed weaker amygdala activation than controls for the contrast of names previously associated with fearful versus neutral faces. Social phobia severity was negatively related to amygdala activation. Moreover, a detailed psychophysiological interaction analysis revealed an inverse correlation between disorder severity and frontolimbic connectivity for the emotional > neutral pseudonames contrast. Our sample included only women. Our results support the theory of a disturbed cortico limbic interplay, even for recently learned emotional stimuli. We discuss the findings with regard to the vigilance-avoidance theory and contrast them to results indicating an oversensitive limbic system in patients with social phobia.

Agile convolutional neural network for pulmonary nodule classification using CT images.

Science.gov (United States)

Zhao, Xinzhuo; Liu, Liyao; Qi, Shouliang; Teng, Yueyang; Li, Jianhua; Qian, Wei

2018-04-01

To distinguish benign from malignant pulmonary nodules using CT images is critical for their precise diagnosis and treatment. A new Agile convolutional neural network (CNN) framework is proposed to conquer the challenges of a small-scale medical image database and the small size of the nodules, and it improves the performance of pulmonary nodule classification using CT images. A hybrid CNN of LeNet and AlexNet is constructed through combining the layer settings of LeNet and the parameter settings of AlexNet. A dataset with 743 CT image nodule samples is built up based on the 1018 CT scans of LIDC to train and evaluate the Agile CNN model. Through adjusting the parameters of the kernel size, learning rate, and other factors, the effect of these parameters on the performance of the CNN model is investigated, and an optimized setting of the CNN is obtained finally. After finely optimizing the settings of the CNN, the estimation accuracy and the area under the curve can reach 0.822 and 0.877, respectively. The accuracy of the CNN is significantly dependent on the kernel size, learning rate, training batch size, dropout, and weight initializations. The best performance is achieved when the kernel size is set to [Formula: see text], the learning rate is 0.005, the batch size is 32, and dropout and Gaussian initialization are used. This competitive performance demonstrates that our proposed CNN framework and the optimization strategy of the CNN parameters are suitable for pulmonary nodule classification characterized by small medical datasets and small targets. The classification model might help diagnose and treat pulmonary nodules effectively.

Supervised learning of probability distributions by neural networks

Science.gov (United States)

Baum, Eric B.; Wilczek, Frank

1988-01-01

Supervised learning algorithms for feedforward neural networks are investigated analytically. The back-propagation algorithm described by Werbos (1974), Parker (1985), and Rumelhart et al. (1986) is generalized by redefining the values of the input and output neurons as probabilities. The synaptic weights are then varied to follow gradients in the logarithm of likelihood rather than in the error. This modification is shown to provide a more rigorous theoretical basis for the algorithm and to permit more accurate predictions. A typical application involving a medical-diagnosis expert system is discussed.

Deep Learning Applications for Predicting Pharmacological Properties of Drugs and Drug Repurposing Using Transcriptomic Data.

Science.gov (United States)

Aliper, Alexander; Plis, Sergey; Artemov, Artem; Ulloa, Alvaro; Mamoshina, Polina; Zhavoronkov, Alex

2016-07-05

Deep learning is rapidly advancing many areas of science and technology with multiple success stories in image, text, voice and video recognition, robotics, and autonomous driving. In this paper we demonstrate how deep neural networks (DNN) trained on large transcriptional response data sets can classify various drugs to therapeutic categories solely based on their transcriptional profiles. We used the perturbation samples of 678 drugs across A549, MCF-7, and PC-3 cell lines from the LINCS Project and linked those to 12 therapeutic use categories derived from MeSH. To train the DNN, we utilized both gene level transcriptomic data and transcriptomic data processed using a pathway activation scoring algorithm, for a pooled data set of samples perturbed with different concentrations of the drug for 6 and 24 hours. In both pathway and gene level classification, DNN achieved high classification accuracy and convincingly outperformed the support vector machine (SVM) model on every multiclass classification problem, however, models based on pathway level data performed significantly better. For the first time we demonstrate a deep learning neural net trained on transcriptomic data to recognize pharmacological properties of multiple drugs across different biological systems and conditions. We also propose using deep neural net confusion matrices for drug repositioning. This work is a proof of principle for applying deep learning to drug discovery and development.

Reinforcement Learning of Linking and Tracing Contours in Recurrent Neural Networks

Science.gov (United States)

Brosch, Tobias; Neumann, Heiko; Roelfsema, Pieter R.

2015-01-01

The processing of a visual stimulus can be subdivided into a number of stages. Upon stimulus presentation there is an early phase of feedforward processing where the visual information is propagated from lower to higher visual areas for the extraction of basic and complex stimulus features. This is followed by a later phase where horizontal connections within areas and feedback connections from higher areas back to lower areas come into play. In this later phase, image elements that are behaviorally relevant are grouped by Gestalt grouping rules and are labeled in the cortex with enhanced neuronal activity (object-based attention in psychology). Recent neurophysiological studies revealed that reward-based learning influences these recurrent grouping processes, but it is not well understood how rewards train recurrent circuits for perceptual organization. This paper examines the mechanisms for reward-based learning of new grouping rules. We derive a learning rule that can explain how rewards influence the information flow through feedforward, horizontal and feedback connections. We illustrate the efficiency with two tasks that have been used to study the neuronal correlates of perceptual organization in early visual cortex. The first task is called contour-integration and demands the integration of collinear contour elements into an elongated curve. We show how reward-based learning causes an enhancement of the representation of the to-be-grouped elements at early levels of a recurrent neural network, just as is observed in the visual cortex of monkeys. The second task is curve-tracing where the aim is to determine the endpoint of an elongated curve composed of connected image elements. If trained with the new learning rule, neural networks learn to propagate enhanced activity over the curve, in accordance with neurophysiological data. We close the paper with a number of model predictions that can be tested in future neurophysiological and computational studies

Feature Extraction and Fusion Using Deep Convolutional Neural Networks for Face Detection

Directory of Open Access Journals (Sweden)

Xiaojun Lu

2017-01-01

Full Text Available This paper proposes a method that uses feature fusion to represent images better for face detection after feature extraction by deep convolutional neural network (DCNN. First, with Clarifai net and VGG Net-D (16 layers, we learn features from data, respectively; then we fuse features extracted from the two nets. To obtain more compact feature representation and mitigate computation complexity, we reduce the dimension of the fused features by PCA. Finally, we conduct face classification by SVM classifier for binary classification. In particular, we exploit offset max-pooling to extract features with sliding window densely, which leads to better matches of faces and detection windows; thus the detection result is more accurate. Experimental results show that our method can detect faces with severe occlusion and large variations in pose and scale. In particular, our method achieves 89.24% recall rate on FDDB and 97.19% average precision on AFW.

Neural substrates underlying stimulation-enhanced motor skill learning after stroke.

Science.gov (United States)

Lefebvre, Stéphanie; Dricot, Laurence; Laloux, Patrice; Gradkowski, Wojciech; Desfontaines, Philippe; Evrard, Frédéric; Peeters, André; Jamart, Jacques; Vandermeeren, Yves

2015-01-01

Motor skill learning is one of the key components of motor function recovery after stroke, especially recovery driven by neurorehabilitation. Transcranial direct current stimulation can enhance neurorehabilitation and motor skill learning in stroke patients. However, the neural mechanisms underlying the retention of stimulation-enhanced motor skill learning involving a paretic upper limb have not been resolved. These neural substrates were explored by means of functional magnetic resonance imaging. Nineteen chronic hemiparetic stroke patients participated in a double-blind, cross-over randomized, sham-controlled experiment with two series. Each series consisted of two sessions: (i) an intervention session during which dual transcranial direct current stimulation or sham was applied during motor skill learning with the paretic upper limb; and (ii) an imaging session 1 week later, during which the patients performed the learned motor skill. The motor skill learning task, called the 'circuit game', involves a speed/accuracy trade-off and consists of moving a pointer controlled by a computer mouse along a complex circuit as quickly and accurately as possible. Relative to the sham series, dual transcranial direct current stimulation applied bilaterally over the primary motor cortex during motor skill learning with the paretic upper limb resulted in (i) enhanced online motor skill learning; (ii) enhanced 1-week retention; and (iii) superior transfer of performance improvement to an untrained task. The 1-week retention's enhancement driven by the intervention was associated with a trend towards normalization of the brain activation pattern during performance of the learned motor skill relative to the sham series. A similar trend towards normalization relative to sham was observed during performance of a simple, untrained task without a speed/accuracy constraint, despite a lack of behavioural difference between the dual transcranial direct current stimulation and sham

Modeling the behavioral substrates of associate learning and memory - Adaptive neural models

Science.gov (United States)

Lee, Chuen-Chien

1991-01-01

Three adaptive single-neuron models based on neural analogies of behavior modification episodes are proposed, which attempt to bridge the gap between psychology and neurophysiology. The proposed models capture the predictive nature of Pavlovian conditioning, which is essential to the theory of adaptive/learning systems. The models learn to anticipate the occurrence of a conditioned response before the presence of a reinforcing stimulus when training is complete. Furthermore, each model can find the most nonredundant and earliest predictor of reinforcement. The behavior of the models accounts for several aspects of basic animal learning phenomena in Pavlovian conditioning beyond previous related models. Computer simulations show how well the models fit empirical data from various animal learning paradigms.

Learning Orthographic Structure With Sequential Generative Neural Networks.

Science.gov (United States)

Testolin, Alberto; Stoianov, Ivilin; Sperduti, Alessandro; Zorzi, Marco

2016-04-01

Learning the structure of event sequences is a ubiquitous problem in cognition and particularly in language. One possible solution is to learn a probabilistic generative model of sequences that allows making predictions about upcoming events. Though appealing from a neurobiological standpoint, this approach is typically not pursued in connectionist modeling. Here, we investigated a sequential version of the restricted Boltzmann machine (RBM), a stochastic recurrent neural network that extracts high-order structure from sensory data through unsupervised generative learning and can encode contextual information in the form of internal, distributed representations. We assessed whether this type of network can extract the orthographic structure of English monosyllables by learning a generative model of the letter sequences forming a word training corpus. We show that the network learned an accurate probabilistic model of English graphotactics, which can be used to make predictions about the letter following a given context as well as to autonomously generate high-quality pseudowords. The model was compared to an extended version of simple recurrent networks, augmented with a stochastic process that allows autonomous generation of sequences, and to non-connectionist probabilistic models (n-grams and hidden Markov models). We conclude that sequential RBMs and stochastic simple recurrent networks are promising candidates for modeling cognition in the temporal domain. Copyright © 2015 Cognitive Science Society, Inc.

Self-teaching neural network learns difficult reactor control problem

International Nuclear Information System (INIS)

Jouse, W.C.

1989-01-01

A self-teaching neural network used as an adaptive controller quickly learns to control an unstable reactor configuration. The network models the behavior of a human operator. It is trained by allowing it to operate the reactivity control impulsively. It is punished whenever either the power or fuel temperature stray outside technical limits. Using a simple paradigm, the network constructs an internal representation of the punishment and of the reactor system. The reactor is constrained to small power orbits

Learning Errors by Radial Basis Function Neural Networks and Regularization Networks

Czech Academy of Sciences Publication Activity Database

Neruda, Roman; Vidnerová, Petra

2009-01-01

Roč. 1, č. 2 (2009), s. 49-57 ISSN 2005-4262 R&D Projects: GA MŠk(CZ) 1M0567 Institutional research plan: CEZ:AV0Z10300504 Keywords : neural network * RBF networks * regularization * learning Subject RIV: IN - Informatics, Computer Science http://www.sersc.org/journals/IJGDC/vol2_no1/5.pdf

Generalization error analysis: deep convolutional neural network in mammography

Science.gov (United States)

Richter, Caleb D.; Samala, Ravi K.; Chan, Heang-Ping; Hadjiiski, Lubomir; Cha, Kenny

2018-02-01

We conducted a study to gain understanding of the generalizability of deep convolutional neural networks (DCNNs) given their inherent capability to memorize data. We examined empirically a specific DCNN trained for classification of masses on mammograms. Using a data set of 2,454 lesions from 2,242 mammographic views, a DCNN was trained to classify masses into malignant and benign classes using transfer learning from ImageNet LSVRC-2010. We performed experiments with varying amounts of label corruption and types of pixel randomization to analyze the generalization error for the DCNN. Performance was evaluated using the area under the receiver operating characteristic curve (AUC) with an N-fold cross validation. Comparisons were made between the convergence times, the inference AUCs for both the training set and the test set of the original image patches without corruption, and the root-mean-squared difference (RMSD) in the layer weights of the DCNN trained with different amounts and methods of corruption. Our experiments observed trends which revealed that the DCNN overfitted by memorizing corrupted data. More importantly, this study improved our understanding of DCNN weight updates when learning new patterns or new labels. Although we used a specific classification task with the ImageNet as example, similar methods may be useful for analysis of the DCNN learning processes, especially those that employ transfer learning for medical image analysis where sample size is limited and overfitting risk is high.

Accelerating learning of neural networks with conjugate gradients for nuclear power plant applications

International Nuclear Information System (INIS)

Reifman, J.; Vitela, J.E.

1994-01-01

The method of conjugate gradients is used to expedite the learning process of feedforward multilayer artificial neural networks and to systematically update both the learning parameter and the momentum parameter at each training cycle. The mechanism for the occurrence of premature saturation of the network nodes observed with the back propagation algorithm is described, suggestions are made to eliminate this undesirable phenomenon, and the reason by which this phenomenon is precluded in the method of conjugate gradients is presented. The proposed method is compared with the standard back propagation algorithm in the training of neural networks to classify transient events in neural power plants simulated by the Midland Nuclear Power Plant Unit 2 simulator. The comparison results indicate that the rate of convergence of the proposed method is much greater than the standard back propagation, that it reduces both the number of training cycles and the CPU time, and that it is less sensitive to the choice of initial weights. The advantages of the method are more noticeable and important for problems where the network architecture consists of a large number of nodes, the training database is large, and a tight convergence criterion is desired

Biologically based neural circuit modelling for the study of fear learning and extinction

Science.gov (United States)

Nair, Satish S.; Paré, Denis; Vicentic, Aleksandra

2016-11-01

The neuronal systems that promote protective defensive behaviours have been studied extensively using Pavlovian conditioning. In this paradigm, an initially neutral-conditioned stimulus is paired with an aversive unconditioned stimulus leading the subjects to display behavioural signs of fear. Decades of research into the neural bases of this simple behavioural paradigm uncovered that the amygdala, a complex structure comprised of several interconnected nuclei, is an essential part of the neural circuits required for the acquisition, consolidation and expression of fear memory. However, emerging evidence from the confluence of electrophysiological, tract tracing, imaging, molecular, optogenetic and chemogenetic methodologies, reveals that fear learning is mediated by multiple connections between several amygdala nuclei and their distributed targets, dynamical changes in plasticity in local circuit elements as well as neuromodulatory mechanisms that promote synaptic plasticity. To uncover these complex relations and analyse multi-modal data sets acquired from these studies, we argue that biologically realistic computational modelling, in conjunction with experiments, offers an opportunity to advance our understanding of the neural circuit mechanisms of fear learning and to address how their dysfunction may lead to maladaptive fear responses in mental disorders.

Getting to Net Zero

Energy Technology Data Exchange (ETDEWEB)

2016-09-01

The technology necessary to build net zero energy buildings (NZEBs) is ready and available today, however, building to net zero energy performance levels can be challenging. Energy efficiency measures, onsite energy generation resources, load matching and grid interaction, climatic factors, and local policies vary from location to location and require unique methods of constructing NZEBs. It is recommended that Components start looking into how to construct and operate NZEBs now as there is a learning curve to net zero construction and FY 2020 is just around the corner.

Pro NET Best Practices

CERN Document Server

Ritchie, Stephen D

2011-01-01

Pro .NET Best Practices is a practical reference to the best practices that you can apply to your .NET projects today. You will learn standards, techniques, and conventions that are sharply focused, realistic and helpful for achieving results, steering clear of unproven, idealistic, and impractical recommendations. Pro .NET Best Practices covers a broad range of practices and principles that development experts agree are the right ways to develop software, which includes continuous integration, automated testing, automated deployment, and code analysis. Whether the solution is from a free and

Neural signals of vicarious extinction learning.

Science.gov (United States)

Golkar, Armita; Haaker, Jan; Selbing, Ida; Olsson, Andreas

2016-10-01

Social transmission of both threat and safety is ubiquitous, but little is known about the neural circuitry underlying vicarious safety learning. This is surprising given that these processes are critical to flexibly adapt to a changeable environment. To address how the expression of previously learned fears can be modified by the transmission of social information, two conditioned stimuli (CS + s) were paired with shock and the third was not. During extinction, we held constant the amount of direct, non-reinforced, exposure to the CSs (i.e. direct extinction), and critically varied whether another individual-acting as a demonstrator-experienced safety (CS + vic safety) or aversive reinforcement (CS + vic reinf). During extinction, ventromedial prefrontal cortex (vmPFC) responses to the CS + vic reinf increased but decreased to the CS + vic safety This pattern of vmPFC activity was reversed during a subsequent fear reinstatement test, suggesting a temporal shift in the involvement of the vmPFC. Moreover, only the CS + vic reinf association recovered. Our data suggest that vicarious extinction prevents the return of conditioned fear responses, and that this efficacy is reflected by diminished vmPFC involvement during extinction learning. The present findings may have important implications for understanding how social information influences the persistence of fear memories in individuals suffering from emotional disorders. © The Author (2016). Published by Oxford University Press. For Permissions, please email: journals.permissions@oup.com.

Multimodal Feature Learning for Video Captioning

Directory of Open Access Journals (Sweden)

Sujin Lee

2018-01-01

Full Text Available Video captioning refers to the task of generating a natural language sentence that explains the content of the input video clips. This study proposes a deep neural network model for effective video captioning. Apart from visual features, the proposed model learns additionally semantic features that describe the video content effectively. In our model, visual features of the input video are extracted using convolutional neural networks such as C3D and ResNet, while semantic features are obtained using recurrent neural networks such as LSTM. In addition, our model includes an attention-based caption generation network to generate the correct natural language captions based on the multimodal video feature sequences. Various experiments, conducted with the two large benchmark datasets, Microsoft Video Description (MSVD and Microsoft Research Video-to-Text (MSR-VTT, demonstrate the performance of the proposed model.

Projection decomposition algorithm for dual-energy computed tomography via deep neural network.

Science.gov (United States)

Xu, Yifu; Yan, Bin; Chen, Jian; Zeng, Lei; Li, Lei

2018-03-15

Dual-energy computed tomography (DECT) has been widely used to improve identification of substances from different spectral information. Decomposition of the mixed test samples into two materials relies on a well-calibrated material decomposition function. This work aims to establish and validate a data-driven algorithm for estimation of the decomposition function. A deep neural network (DNN) consisting of two sub-nets is proposed to solve the projection decomposition problem. The compressing sub-net, substantially a stack auto-encoder (SAE), learns a compact representation of energy spectrum. The decomposing sub-net with a two-layer structure fits the nonlinear transform between energy projection and basic material thickness. The proposed DNN not only delivers image with lower standard deviation and higher quality in both simulated and real data, and also yields the best performance in cases mixed with photon noise. Moreover, DNN costs only 0.4 s to generate a decomposition solution of 360 × 512 size scale, which is about 200 times faster than the competing algorithms. The DNN model is applicable to the decomposition tasks with different dual energies. Experimental results demonstrated the strong function fitting ability of DNN. Thus, the Deep learning paradigm provides a promising approach to solve the nonlinear problem in DECT.

Learning to read words in a new language shapes the neural organization of the prior languages.

Science.gov (United States)

Mei, Leilei; Xue, Gui; Lu, Zhong-Lin; Chen, Chuansheng; Zhang, Mingxia; He, Qinghua; Wei, Miao; Dong, Qi

2014-12-01

Learning a new language entails interactions with one׳s prior language(s). Much research has shown how native language affects the cognitive and neural mechanisms of a new language, but little is known about whether and how learning a new language shapes the neural mechanisms of prior language(s). In two experiments in the current study, we used an artificial language training paradigm in combination with an fMRI to examine (1) the effects of different linguistic components (phonology and semantics) of a new language on the neural process of prior languages (i.e., native and second languages), and (2) whether such effects were modulated by the proficiency level in the new language. Results of Experiment 1 showed that when the training in a new language involved semantics (as opposed to only visual forms and phonology), neural activity during word reading in the native language (Chinese) was reduced in several reading-related regions, including the left pars opercularis, pars triangularis, bilateral inferior temporal gyrus, fusiform gyrus, and inferior occipital gyrus. Results of Experiment 2 replicated the results of Experiment 1 and further found that semantic training also affected neural activity during word reading in the subjects׳ second language (English). Furthermore, we found that the effects of the new language were modulated by the subjects׳ proficiency level in the new language. These results provide critical imaging evidence for the influence of learning to read words in a new language on word reading in native and second languages. Copyright © 2014 Elsevier Ltd. All rights reserved.

Identifying beneficial task relations for multi-task learning in deep neural networks

DEFF Research Database (Denmark)

Bingel, Joachim; Søgaard, Anders

2017-01-01

Multi-task learning (MTL) in deep neural networks for NLP has recently received increasing interest due to some compelling benefits, including its potential to efficiently regularize models and to reduce the need for labeled data. While it has brought significant improvements in a number of NLP...

Using c-Jun to identify fear extinction learning-specific patterns of neural activity that are affected by single prolonged stress.

Science.gov (United States)

Knox, Dayan; Stanfield, Briana R; Staib, Jennifer M; David, Nina P; DePietro, Thomas; Chamness, Marisa; Schneider, Elizabeth K; Keller, Samantha M; Lawless, Caroline

2018-04-02

Neural circuits via which stress leads to disruptions in fear extinction is often explored in animal stress models. Using the single prolonged stress (SPS) model of post traumatic stress disorder and the immediate early gene (IEG) c-Fos as a measure of neural activity, we previously identified patterns of neural activity through which SPS disrupts extinction retention. However, none of these stress effects were specific to fear or extinction learning and memory. C-Jun is another IEG that is sometimes regulated in a different manner to c-Fos and could be used to identify emotional learning/memory specific patterns of neural activity that are sensitive to SPS. Animals were either fear conditioned (CS-fear) or presented with CSs only (CS-only) then subjected to extinction training and testing. C-Jun was then assayed within neural substrates critical for extinction memory. Inhibited c-Jun levels in the hippocampus (Hipp) and enhanced functional connectivity between the ventromedial prefrontal cortex (vmPFC) and basolateral amygdala (BLA) during extinction training was disrupted by SPS in the CS-fear group only. As a result, these effects were specific to emotional learning/memory. SPS also disrupted inhibited Hipp c-Jun levels, enhanced BLA c-Jun levels, and altered functional connectivity among the vmPFC, BLA, and Hipp during extinction testing in SPS rats in the CS-fear and CS-only groups. As a result, these effects were not specific to emotional learning/memory. Our findings suggest that SPS disrupts neural activity specific to extinction memory, but may also disrupt the retention of fear extinction by mechanisms that do not involve emotional learning/memory. Copyright © 2017 Elsevier B.V. All rights reserved.

Fuzzy-neural network in the automatic detection and volumetry of the spleen on spiral CT scans

International Nuclear Information System (INIS)

Heitmann, K.R.; Mainz Univ.; Rueckert, S.; Heussel, C.P.; Thelen, M.; Kauczor, H.U.; Uthmann, T.

2000-01-01

Purpose: To assess spleen segmentation and volumetry in spiral CT scans with and without pathological changes of splenic tissue. Methods: The image analysis software HYBRIKON is based on region growing, self-organized neural nets, and fuzzy-anatomic rules. The neural nets were trained with spiral CT data from 10 patients, not used in the following evaluation on spiral CT scans from 19 patients. An experienced radiologist verified the results. The true positive and false positive areas were compared in terms to the areas marked by the radiologist. The results were compared with a standard thresholding method. Results: The neural nets achieved a higher accuracy than the thresholding method. Correlation coefficient of the fuzzy-neural nets: 0.99 (thresholding: 0.63). Mean true positive rate: 90% (thresholding: 75%), mean false positive rate: 5% (thresholding>100%). Pitfalls were caused by accessory spleens, extreme changes in the morphology (tumors, metastases, cysts), and parasplenic masses. Conclusions: Self-organizing neural nets combined with fuzzy rules are ready for use in the automatic detection and volumetry of the spleen in spiral CT scans. (orig.) [de

Relabeling exchange method (REM) for learning in neural networks

Science.gov (United States)

Wu, Wen; Mammone, Richard J.

1994-02-01

The supervised training of neural networks require the use of output labels which are usually arbitrarily assigned. In this paper it is shown that there is a significant difference in the rms error of learning when `optimal' label assignment schemes are used. We have investigated two efficient random search algorithms to solve the relabeling problem: the simulated annealing and the genetic algorithm. However, we found them to be computationally expensive. Therefore we shall introduce a new heuristic algorithm called the Relabeling Exchange Method (REM) which is computationally more attractive and produces optimal performance. REM has been used to organize the optimal structure for multi-layered perceptrons and neural tree networks. The method is a general one and can be implemented as a modification to standard training algorithms. The motivation of the new relabeling strategy is based on the present interpretation of dyslexia as an encoding problem.

Statistical physics, neural networks, brain studies

International Nuclear Information System (INIS)

Toulouse, G.

1999-01-01

An overview of some aspects of a vast domain, located at the crossroads of physics, biology and computer science is presented: (1) During the last fifteen years, physicists advancing along various pathways have come into contact with biology (computational neurosciences) and engineering (formal neural nets). (2) This move may actually be viewed as one component in a larger picture. A prominent trend of recent years, observable over many countries, has been the establishment of interdisciplinary centers devoted to the study of: cognitive sciences; natural and artificial intelligence; brain, mind and behaviour; perception and action; learning and memory; robotics; man-machine communication, etc. What are the promising lines of development? What opportunities for physicists? An attempt will be made to address such questions and related issues

A mathematical analysis of the effects of Hebbian learning rules on the dynamics and structure of discrete-time random recurrent neural networks.

Science.gov (United States)

Siri, Benoît; Berry, Hugues; Cessac, Bruno; Delord, Bruno; Quoy, Mathias

2008-12-01

We present a mathematical analysis of the effects of Hebbian learning in random recurrent neural networks, with a generic Hebbian learning rule, including passive forgetting and different timescales, for neuronal activity and learning dynamics. Previous numerical work has reported that Hebbian learning drives the system from chaos to a steady state through a sequence of bifurcations. Here, we interpret these results mathematically and show that these effects, involving a complex coupling between neuronal dynamics and synaptic graph structure, can be analyzed using Jacobian matrices, which introduce both a structural and a dynamical point of view on neural network evolution. Furthermore, we show that sensitivity to a learned pattern is maximal when the largest Lyapunov exponent is close to 0. We discuss how neural networks may take advantage of this regime of high functional interest.

From phonemes to images : levels of representation in a recurrent neural model of visually-grounded language learning

NARCIS (Netherlands)

Gelderloos, L.J.; Chrupala, Grzegorz

2016-01-01

We present a model of visually-grounded language learning based on stacked gated recurrent neural networks which learns to predict visual features given an image description in the form of a sequence of phonemes. The learning task resembles that faced by human language learners who need to discover

Neural Pattern Similarity in the Left IFG and Fusiform Is Associated with Novel Word Learning

Directory of Open Access Journals (Sweden)

Jing Qu

2017-08-01

Full Text Available Previous studies have revealed that greater neural pattern similarity across repetitions is associated with better subsequent memory. In this study, we used an artificial language training paradigm and representational similarity analysis to examine whether neural pattern similarity across repetitions before training was associated with post-training behavioral performance. Twenty-four native Chinese speakers were trained to learn a logographic artificial language for 12 days and behavioral performance was recorded using the word naming and picture naming tasks. Participants were scanned while performing a passive viewing task before training, after 4-day training and after 12-day training. Results showed that pattern similarity in the left pars opercularis (PO and fusiform gyrus (FG before training was negatively associated with reaction time (RT in both word naming and picture naming tasks after training. These results suggest that neural pattern similarity is an effective neurofunctional predictor of novel word learning in addition to word memory.

Neural Pattern Similarity in the Left IFG and Fusiform Is Associated with Novel Word Learning

Science.gov (United States)

Qu, Jing; Qian, Liu; Chen, Chuansheng; Xue, Gui; Li, Huiling; Xie, Peng; Mei, Leilei

2017-01-01

Previous studies have revealed that greater neural pattern similarity across repetitions is associated with better subsequent memory. In this study, we used an artificial language training paradigm and representational similarity analysis to examine whether neural pattern similarity across repetitions before training was associated with post-training behavioral performance. Twenty-four native Chinese speakers were trained to learn a logographic artificial language for 12 days and behavioral performance was recorded using the word naming and picture naming tasks. Participants were scanned while performing a passive viewing task before training, after 4-day training and after 12-day training. Results showed that pattern similarity in the left pars opercularis (PO) and fusiform gyrus (FG) before training was negatively associated with reaction time (RT) in both word naming and picture naming tasks after training. These results suggest that neural pattern similarity is an effective neurofunctional predictor of novel word learning in addition to word memory. PMID:28878640

Application of artificial neural network with extreme learning machine for economic growth estimation

Science.gov (United States)

Milačić, Ljubiša; Jović, Srđan; Vujović, Tanja; Miljković, Jovica

2017-01-01

The purpose of this research is to develop and apply the artificial neural network (ANN) with extreme learning machine (ELM) to forecast gross domestic product (GDP) growth rate. The economic growth forecasting was analyzed based on agriculture, manufacturing, industry and services value added in GDP. The results were compared with ANN with back propagation (BP) learning approach since BP could be considered as conventional learning methodology. The reliability of the computational models was accessed based on simulation results and using several statistical indicators. Based on results, it was shown that ANN with ELM learning methodology can be applied effectively in applications of GDP forecasting.

The principles of artificial neural network information processing

International Nuclear Information System (INIS)

Dai, Ru-Wei

1993-01-01

In this article, the basic structure of an artificial neuron is first introduced. In addition, principles of artificial neural network as well as several important artificial neural models such as Perceptron, Back propagation model, Hopfield net, and ART model are briefly discussed and analyzed. Finally, the application of artificial neural network for Chinese Character Recognition is also given. (author)

The principles of artificial neural network information processing

International Nuclear Information System (INIS)

Dai, Ru-Wei

1993-01-01

In this article, the basic structure of an artificial neuron is first introduced. In addition, principles of artificial neural network as well as several important artificial neural models such as perception, back propagation model, Hopfield net, and ART model are briefly discussed and analyzed. Finally the application of artificial neural network for Chinese character recognition is also given. (author)

Differences between Neural Activity in Prefrontal Cortex and Striatum during Learning of Novel Abstract Categories

OpenAIRE

Antzoulatos, Evan G.; Miller, Earl K.

2011-01-01

Learning to classify diverse experiences into meaningful groups, like categories, is fundamental to normal cognition. To understand its neural basis, we simultaneously recorded from multiple electrodes in the lateral prefrontal cortex and dorsal striatum, two interconnected brain structures critical for learning. Each day, monkeys learned to associate novel, abstract dot-based categories with a right vs. left saccade. Early on, when they could acquire specific stimulus-response associations, ...

Single-shot T2 mapping using overlapping-echo detachment planar imaging and a deep convolutional neural network.

Science.gov (United States)

Cai, Congbo; Wang, Chao; Zeng, Yiqing; Cai, Shuhui; Liang, Dong; Wu, Yawen; Chen, Zhong; Ding, Xinghao; Zhong, Jianhui

2018-04-24

An end-to-end deep convolutional neural network (CNN) based on deep residual network (ResNet) was proposed to efficiently reconstruct reliable T 2 mapping from single-shot overlapping-echo detachment (OLED) planar imaging. The training dataset was obtained from simulations that were carried out on SPROM (Simulation with PRoduct Operator Matrix) software developed by our group. The relationship between the original OLED image containing two echo signals and the corresponding T 2 mapping was learned by ResNet training. After the ResNet was trained, it was applied to reconstruct the T 2 mapping from simulation and in vivo human brain data. Although the ResNet was trained entirely on simulated data, the trained network was generalized well to real human brain data. The results from simulation and in vivo human brain experiments show that the proposed method significantly outperforms the echo-detachment-based method. Reliable T 2 mapping with higher accuracy is achieved within 30 ms after the network has been trained, while the echo-detachment-based OLED reconstruction method took approximately 2 min. The proposed method will facilitate real-time dynamic and quantitative MR imaging via OLED sequence, and deep convolutional neural network has the potential to reconstruct maps from complex MRI sequences efficiently. © 2018 International Society for Magnetic Resonance in Medicine.

Dynamics of neural cryptography.

Science.gov (United States)

Ruttor, Andreas; Kinzel, Wolfgang; Kanter, Ido

2007-05-01

Synchronization of neural networks has been used for public channel protocols in cryptography. In the case of tree parity machines the dynamics of both bidirectional synchronization and unidirectional learning is driven by attractive and repulsive stochastic forces. Thus it can be described well by a random walk model for the overlap between participating neural networks. For that purpose transition probabilities and scaling laws for the step sizes are derived analytically. Both these calculations as well as numerical simulations show that bidirectional interaction leads to full synchronization on average. In contrast, successful learning is only possible by means of fluctuations. Consequently, synchronization is much faster than learning, which is essential for the security of the neural key-exchange protocol. However, this qualitative difference between bidirectional and unidirectional interaction vanishes if tree parity machines with more than three hidden units are used, so that those neural networks are not suitable for neural cryptography. In addition, the effective number of keys which can be generated by the neural key-exchange protocol is calculated using the entropy of the weight distribution. As this quantity increases exponentially with the system size, brute-force attacks on neural cryptography can easily be made unfeasible.

Dynamics of neural cryptography

International Nuclear Information System (INIS)

Ruttor, Andreas; Kinzel, Wolfgang; Kanter, Ido

2007-01-01

Synchronization of neural networks has been used for public channel protocols in cryptography. In the case of tree parity machines the dynamics of both bidirectional synchronization and unidirectional learning is driven by attractive and repulsive stochastic forces. Thus it can be described well by a random walk model for the overlap between participating neural networks. For that purpose transition probabilities and scaling laws for the step sizes are derived analytically. Both these calculations as well as numerical simulations show that bidirectional interaction leads to full synchronization on average. In contrast, successful learning is only possible by means of fluctuations. Consequently, synchronization is much faster than learning, which is essential for the security of the neural key-exchange protocol. However, this qualitative difference between bidirectional and unidirectional interaction vanishes if tree parity machines with more than three hidden units are used, so that those neural networks are not suitable for neural cryptography. In addition, the effective number of keys which can be generated by the neural key-exchange protocol is calculated using the entropy of the weight distribution. As this quantity increases exponentially with the system size, brute-force attacks on neural cryptography can easily be made unfeasible

Dynamics of neural cryptography

Science.gov (United States)

Ruttor, Andreas; Kinzel, Wolfgang; Kanter, Ido

2007-05-01

Synchronization of neural networks has been used for public channel protocols in cryptography. In the case of tree parity machines the dynamics of both bidirectional synchronization and unidirectional learning is driven by attractive and repulsive stochastic forces. Thus it can be described well by a random walk model for the overlap between participating neural networks. For that purpose transition probabilities and scaling laws for the step sizes are derived analytically. Both these calculations as well as numerical simulations show that bidirectional interaction leads to full synchronization on average. In contrast, successful learning is only possible by means of fluctuations. Consequently, synchronization is much faster than learning, which is essential for the security of the neural key-exchange protocol. However, this qualitative difference between bidirectional and unidirectional interaction vanishes if tree parity machines with more than three hidden units are used, so that those neural networks are not suitable for neural cryptography. In addition, the effective number of keys which can be generated by the neural key-exchange protocol is calculated using the entropy of the weight distribution. As this quantity increases exponentially with the system size, brute-force attacks on neural cryptography can easily be made unfeasible.

Evaluation of a deep learning architecture for MR imaging prediction of ATRX in glioma patients

Science.gov (United States)

Korfiatis, Panagiotis; Kline, Timothy L.; Erickson, Bradley J.

2018-02-01

Predicting mutation/loss of alpha-thalassemia/mental retardation syndrome X-linked (ATRX) gene utilizing MR imaging is of high importance since it is a predictor of response and prognosis in brain tumors. In this study, we compare a deep neural network approach based on a residual deep neural network (ResNet) architecture and one based on a classical machine learning approach and evaluate their ability in predicting ATRX mutation status without the need for a distinct tumor segmentation step. We found that the ResNet50 (50 layers) architecture, pre trained on ImageNet data was the best performing model, achieving an accuracy of 0.91 for the test set (classification of a slice as no tumor, ATRX mutated, or mutated) in terms of f1 score in a test set of 35 cases. The SVM classifier achieved 0.63 for differentiating the Flair signal abnormality regions from the test patients based on their mutation status. We report a method that alleviates the need for extensive preprocessing and acts as a proof of concept that deep neural network architectures can be used to predict molecular biomarkers from routine medical images.

Pro ASP.NET 4 in VB 2010

CERN Document Server

MacDonald, Matthew; Freeman, Adam; Szpuszta, Mario; Agarwal, Vidya Vrat

2010-01-01

ASP.NET 4 is the latest version of Microsoft's revolutionary ASP.NET technology. It is the principal standard for creating dynamic web pages on the Windows platform. Pro ASP.NET 4 in VB 2010 raises the bar for high-quality, practical advice on learning and deploying Microsoft's dynamic web solution. This new edition is updated with everything you need to come to grips with the latest version of ASP.NET, including coverage of ASP.NET MVC, ASP.NET AJAX 4, ASP.NET Dynamic Data, and Silverlight 3. Seasoned .NET professionals Matthew MacDonald and Mario Szpuszta explain how you can get the most fro

A neural fuzzy controller learning by fuzzy error propagation

Science.gov (United States)

Nauck, Detlef; Kruse, Rudolf

1992-01-01

In this paper, we describe a procedure to integrate techniques for the adaptation of membership functions in a linguistic variable based fuzzy control environment by using neural network learning principles. This is an extension to our work. We solve this problem by defining a fuzzy error that is propagated back through the architecture of our fuzzy controller. According to this fuzzy error and the strength of its antecedent each fuzzy rule determines its amount of error. Depending on the current state of the controlled system and the control action derived from the conclusion, each rule tunes the membership functions of its antecedent and its conclusion. By this we get an unsupervised learning technique that enables a fuzzy controller to adapt to a control task by knowing just about the global state and the fuzzy error.

Robots Learn to Recognize Individuals from Imitative Encounters with People and Avatars

Science.gov (United States)

Boucenna, Sofiane; Cohen, David; Meltzoff, Andrew N.; Gaussier, Philippe; Chetouani, Mohamed

2016-02-01

Prior to language, human infants are prolific imitators. Developmental science grounds infant imitation in the neural coding of actions, and highlights the use of imitation for learning from and about people. Here, we used computational modeling and a robot implementation to explore the functional value of action imitation. We report 3 experiments using a mutual imitation task between robots, adults, typically developing children, and children with Autism Spectrum Disorder. We show that a particular learning architecture - specifically one combining artificial neural nets for (i) extraction of visual features, (ii) the robot’s motor internal state, (iii) posture recognition, and (iv) novelty detection - is able to learn from an interactive experience involving mutual imitation. This mutual imitation experience allowed the robot to recognize the interactive agent in a subsequent encounter. These experiments using robots as tools for modeling human cognitive development, based on developmental theory, confirm the promise of developmental robotics. Additionally, findings illustrate how person recognition may emerge through imitative experience, intercorporeal mapping, and statistical learning.

CGBayesNets: conditional Gaussian Bayesian network learning and inference with mixed discrete and continuous data.

Science.gov (United States)

McGeachie, Michael J; Chang, Hsun-Hsien; Weiss, Scott T

2014-06-01

Bayesian Networks (BN) have been a popular predictive modeling formalism in bioinformatics, but their application in modern genomics has been slowed by an inability to cleanly handle domains with mixed discrete and continuous variables. Existing free BN software packages either discretize continuous variables, which can lead to information loss, or do not include inference routines, which makes prediction with the BN impossible. We present CGBayesNets, a BN package focused around prediction of a clinical phenotype from mixed discrete and continuous variables, which fills these gaps. CGBayesNets implements Bayesian likelihood and inference algorithms for the conditional Gaussian Bayesian network (CGBNs) formalism, one appropriate for predicting an outcome of interest from, e.g., multimodal genomic data. We provide four different network learning algorithms, each making a different tradeoff between computational cost and network likelihood. CGBayesNets provides a full suite of functions for model exploration and verification, including cross validation, bootstrapping, and AUC manipulation. We highlight several results obtained previously with CGBayesNets, including predictive models of wood properties from tree genomics, leukemia subtype classification from mixed genomic data, and robust prediction of intensive care unit mortality outcomes from metabolomic profiles. We also provide detailed example analysis on public metabolomic and gene expression datasets. CGBayesNets is implemented in MATLAB and available as MATLAB source code, under an Open Source license and anonymous download at http://www.cgbayesnets.com.

Failing to learn from negative prediction errors: Obesity is associated with alterations in a fundamental neural learning mechanism.

Science.gov (United States)

Mathar, David; Neumann, Jane; Villringer, Arno; Horstmann, Annette

2017-10-01

Prediction errors (PEs) encode the difference between expected and actual action outcomes in the brain via dopaminergic modulation. Integration of these learning signals ensures efficient behavioral adaptation. Obesity has recently been linked to altered dopaminergic fronto-striatal circuits, thus implying impairments in cognitive domains that rely on its integrity. 28 obese and 30 lean human participants performed an implicit stimulus-response learning paradigm inside an fMRI scanner. Computational modeling and psycho-physiological interaction (PPI) analysis was utilized for assessing PE-related learning and associated functional connectivity. We show that human obesity is associated with insufficient incorporation of negative PEs into behavioral adaptation even in a non-food context, suggesting differences in a fundamental neural learning mechanism. Obese subjects were less efficient in using negative PEs to improve implicit learning performance, despite proper coding of PEs in striatum. We further observed lower functional coupling between ventral striatum and supplementary motor area in obese subjects subsequent to negative PEs. Importantly, strength of functional coupling predicted task performance and negative PE utilization. These findings show that obesity is linked to insufficient behavioral adaptation specifically in response to negative PEs, and to associated alterations in function and connectivity within the fronto-striatal system. Recognition of neural differences as a central characteristic of obesity hopefully paves the way to rethink established intervention strategies: Differential behavioral sensitivity to negative and positive PEs should be considered when designing intervention programs. Measures relying on penalization of unwanted behavior may prove less effective in obese subjects than alternative approaches. Copyright © 2017 Elsevier Ltd. All rights reserved.

Deep neural networks show an equivalent and often superior performance to dermatologists in onychomycosis diagnosis: Automatic construction of onychomycosis datasets by region-based convolutional deep neural network.

Directory of Open Access Journals (Sweden)

Seung Seog Han

Full Text Available Although there have been reports of the successful diagnosis of skin disorders using deep learning, unrealistically large clinical image datasets are required for artificial intelligence (AI training. We created datasets of standardized nail images using a region-based convolutional neural network (R-CNN trained to distinguish the nail from the background. We used R-CNN to generate training datasets of 49,567 images, which we then used to fine-tune the ResNet-152 and VGG-19 models. The validation datasets comprised 100 and 194 images from Inje University (B1 and B2 datasets, respectively, 125 images from Hallym University (C dataset, and 939 images from Seoul National University (D dataset. The AI (ensemble model; ResNet-152 + VGG-19 + feedforward neural networks results showed test sensitivity/specificity/ area under the curve values of (96.0 / 94.7 / 0.98, (82.7 / 96.7 / 0.95, (92.3 / 79.3 / 0.93, (87.7 / 69.3 / 0.82 for the B1, B2, C, and D datasets. With a combination of the B1 and C datasets, the AI Youden index was significantly (p = 0.01 higher than that of 42 dermatologists doing the same assessment manually. For B1+C and B2+ D dataset combinations, almost none of the dermatologists performed as well as the AI. By training with a dataset comprising 49,567 images, we achieved a diagnostic accuracy for onychomycosis using deep learning that was superior to that of most of the dermatologists who participated in this study.

Deep neural networks show an equivalent and often superior performance to dermatologists in onychomycosis diagnosis: Automatic construction of onychomycosis datasets by region-based convolutional deep neural network.

Science.gov (United States)

Han, Seung Seog; Park, Gyeong Hun; Lim, Woohyung; Kim, Myoung Shin; Na, Jung Im; Park, Ilwoo; Chang, Sung Eun

2018-01-01

Although there have been reports of the successful diagnosis of skin disorders using deep learning, unrealistically large clinical image datasets are required for artificial intelligence (AI) training. We created datasets of standardized nail images using a region-based convolutional neural network (R-CNN) trained to distinguish the nail from the background. We used R-CNN to generate training datasets of 49,567 images, which we then used to fine-tune the ResNet-152 and VGG-19 models. The validation datasets comprised 100 and 194 images from Inje University (B1 and B2 datasets, respectively), 125 images from Hallym University (C dataset), and 939 images from Seoul National University (D dataset). The AI (ensemble model; ResNet-152 + VGG-19 + feedforward neural networks) results showed test sensitivity/specificity/ area under the curve values of (96.0 / 94.7 / 0.98), (82.7 / 96.7 / 0.95), (92.3 / 79.3 / 0.93), (87.7 / 69.3 / 0.82) for the B1, B2, C, and D datasets. With a combination of the B1 and C datasets, the AI Youden index was significantly (p = 0.01) higher than that of 42 dermatologists doing the same assessment manually. For B1+C and B2+ D dataset combinations, almost none of the dermatologists performed as well as the AI. By training with a dataset comprising 49,567 images, we achieved a diagnostic accuracy for onychomycosis using deep learning that was superior to that of most of the dermatologists who participated in this study.

Neural correlates of contextual cueing are modulated by explicit learning.

Science.gov (United States)

Westerberg, Carmen E; Miller, Brennan B; Reber, Paul J; Cohen, Neal J; Paller, Ken A

2011-10-01

Contextual cueing refers to the facilitated ability to locate a particular visual element in a scene due to prior exposure to the same scene. This facilitation is thought to reflect implicit learning, as it typically occurs without the observer's knowledge that scenes repeat. Unlike most other implicit learning effects, contextual cueing can be impaired following damage to the medial temporal lobe. Here we investigated neural correlates of contextual cueing and explicit scene memory in two participant groups. Only one group was explicitly instructed about scene repetition. Participants viewed a sequence of complex scenes that depicted a landscape with five abstract geometric objects. Superimposed on each object was a letter T or L rotated left or right by 90°. Participants responded according to the target letter (T) orientation. Responses were highly accurate for all scenes. Response speeds were faster for repeated versus novel scenes. The magnitude of this contextual cueing did not differ between the two groups. Also, in both groups repeated scenes yielded reduced hemodynamic activation compared with novel scenes in several regions involved in visual perception and attention, and reductions in some of these areas were correlated with response-time facilitation. In the group given instructions about scene repetition, recognition memory for scenes was superior and was accompanied by medial temporal and more anterior activation. Thus, strategic factors can promote explicit memorization of visual scene information, which appears to engage additional neural processing beyond what is required for implicit learning of object configurations and target locations in a scene. Copyright © 2011 Elsevier Ltd. All rights reserved.

Spiking neural networks for handwritten digit recognition-Supervised learning and network optimization.

Science.gov (United States)

Kulkarni, Shruti R; Rajendran, Bipin

2018-07-01

We demonstrate supervised learning in Spiking Neural Networks (SNNs) for the problem of handwritten digit recognition using the spike triggered Normalized Approximate Descent (NormAD) algorithm. Our network that employs neurons operating at sparse biological spike rates below 300Hz achieves a classification accuracy of 98.17% on the MNIST test database with four times fewer parameters compared to the state-of-the-art. We present several insights from extensive numerical experiments regarding optimization of learning parameters and network configuration to improve its accuracy. We also describe a number of strategies to optimize the SNN for implementation in memory and energy constrained hardware, including approximations in computing the neuronal dynamics and reduced precision in storing the synaptic weights. Experiments reveal that even with 3-bit synaptic weights, the classification accuracy of the designed SNN does not degrade beyond 1% as compared to the floating-point baseline. Further, the proposed SNN, which is trained based on the precise spike timing information outperforms an equivalent non-spiking artificial neural network (ANN) trained using back propagation, especially at low bit precision. Thus, our study shows the potential for realizing efficient neuromorphic systems that use spike based information encoding and learning for real-world applications. Copyright © 2018 Elsevier Ltd. All rights reserved.

Neural dynamics of learning sound-action associations.

Directory of Open Access Journals (Sweden)

Adam McNamara

Full Text Available A motor component is pre-requisite to any communicative act as one must inherently move to communicate. To learn to make a communicative act, the brain must be able to dynamically associate arbitrary percepts to the neural substrate underlying the pre-requisite motor activity. We aimed to investigate whether brain regions involved in complex gestures (ventral pre-motor cortex, Brodmann Area 44 were involved in mediating association between novel abstract auditory stimuli and novel gestural movements. In a functional resonance imaging (fMRI study we asked participants to learn associations between previously unrelated novel sounds and meaningless gestures inside the scanner. We use functional connectivity analysis to eliminate the often present confound of 'strategic covert naming' when dealing with BA44 and to rule out effects of non-specific reductions in signal. Brodmann Area 44, a region incorporating Broca's region showed strong, bilateral, negative correlation of BOLD (blood oxygen level dependent response with learning of sound-action associations during data acquisition. Left-inferior-parietal-lobule (l-IPL and bilateral loci in and around visual area V5, right-orbital-frontal-gyrus, right-hippocampus, left-para-hippocampus, right-head-of-caudate, right-insula and left-lingual-gyrus also showed decreases in BOLD response with learning. Concurrent with these decreases in BOLD response, an increasing connectivity between areas of the imaged network as well as the right-middle-frontal-gyrus with rising learning performance was revealed by a psychophysiological interaction (PPI analysis. The increasing connectivity therefore occurs within an increasingly energy efficient network as learning proceeds. Strongest learning related connectivity between regions was found when analysing BA44 and l-IPL seeds. The results clearly show that BA44 and l-IPL is dynamically involved in linking gesture and sound and therefore provides evidence that one of

Neural nets for massively parallel optimization

Science.gov (United States)

Dixon, Laurence C. W.; Mills, David

1992-07-01

To apply massively parallel processing systems to the solution of large scale optimization problems it is desirable to be able to evaluate any function f(z), z (epsilon) Rn in a parallel manner. The theorem of Cybenko, Hecht Nielsen, Hornik, Stinchcombe and White, and Funahasi shows that this can be achieved by a neural network with one hidden layer. In this paper we address the problem of the number of nodes required in the layer to achieve a given accuracy in the function and gradient values at all points within a given n dimensional interval. The type of activation function needed to obtain nonsingular Hessian matrices is described and a strategy for obtaining accurate minimal networks presented.

Relay Backpropagation for Effective Learning of Deep Convolutional Neural Networks

OpenAIRE

Shen, Li; Lin, Zhouchen; Huang, Qingming

2015-01-01

Learning deeper convolutional neural networks becomes a tendency in recent years. However, many empirical evidences suggest that performance improvement cannot be gained by simply stacking more layers. In this paper, we consider the issue from an information theoretical perspective, and propose a novel method Relay Backpropagation, that encourages the propagation of effective information through the network in training stage. By virtue of the method, we achieved the first place in ILSVRC 2015...

Neural network representation and learning of mappings and their derivatives

Science.gov (United States)

White, Halbert; Hornik, Kurt; Stinchcombe, Maxwell; Gallant, A. Ronald

1991-01-01

Discussed here are recent theorems proving that artificial neural networks are capable of approximating an arbitrary mapping and its derivatives as accurately as desired. This fact forms the basis for further results establishing the learnability of the desired approximations, using results from non-parametric statistics. These results have potential applications in robotics, chaotic dynamics, control, and sensitivity analysis. An example involving learning the transfer function and its derivatives for a chaotic map is discussed.

Illusory Motion Reproduced by Deep Neural Networks Trained for Prediction.

Science.gov (United States)

Watanabe, Eiji; Kitaoka, Akiyoshi; Sakamoto, Kiwako; Yasugi, Masaki; Tanaka, Kenta

2018-01-01

The cerebral cortex predicts visual motion to adapt human behavior to surrounding objects moving in real time. Although the underlying mechanisms are still unknown, predictive coding is one of the leading theories. Predictive coding assumes that the brain's internal models (which are acquired through learning) predict the visual world at all times and that errors between the prediction and the actual sensory input further refine the internal models. In the past year, deep neural networks based on predictive coding were reported for a video prediction machine called PredNet. If the theory substantially reproduces the visual information processing of the cerebral cortex, then PredNet can be expected to represent the human visual perception of motion. In this study, PredNet was trained with natural scene videos of the self-motion of the viewer, and the motion prediction ability of the obtained computer model was verified using unlearned videos. We found that the computer model accurately predicted the magnitude and direction of motion of a rotating propeller in unlearned videos. Surprisingly, it also represented the rotational motion for illusion images that were not moving physically, much like human visual perception. While the trained network accurately reproduced the direction of illusory rotation, it did not detect motion components in negative control pictures wherein people do not perceive illusory motion. This research supports the exciting idea that the mechanism assumed by the predictive coding theory is one of basis of motion illusion generation. Using sensory illusions as indicators of human perception, deep neural networks are expected to contribute significantly to the development of brain research.

Stroking the Net Whale: A Constructivist Grounded Theory of Self-Regulated Learning in Virtual Social Spaces

Science.gov (United States)

Kasperiuniene, Judita; Zydziunaite, Vilma; Eriksson, Malin

2017-01-01

This qualitative study explored the self-regulated learning (SRL) of teachers and their students in virtual social spaces. The processes of SRL were analyzed from 24 semi-structured individual interviews with professors, instructors and their students from five Lithuanian universities. A core category stroking the net whale showed the process of…

Prototype-Incorporated Emotional Neural Network.

Science.gov (United States)

Oyedotun, Oyebade K; Khashman, Adnan

2017-08-15

Artificial neural networks (ANNs) aim to simulate the biological neural activities. Interestingly, many ''engineering'' prospects in ANN have relied on motivations from cognition and psychology studies. So far, two important learning theories that have been subject of active research are the prototype and adaptive learning theories. The learning rules employed for ANNs can be related to adaptive learning theory, where several examples of the different classes in a task are supplied to the network for adjusting internal parameters. Conversely, the prototype-learning theory uses prototypes (representative examples); usually, one prototype per class of the different classes contained in the task. These prototypes are supplied for systematic matching with new examples so that class association can be achieved. In this paper, we propose and implement a novel neural network algorithm based on modifying the emotional neural network (EmNN) model to unify the prototype- and adaptive-learning theories. We refer to our new model as ``prototype-incorporated EmNN''. Furthermore, we apply the proposed model to two real-life challenging tasks, namely, static hand-gesture recognition and face recognition, and compare the result to those obtained using the popular back-propagation neural network (BPNN), emotional BPNN (EmNN), deep networks, an exemplar classification model, and k-nearest neighbor.

Control of 12-Cylinder Camless Engine with Neural Networks

OpenAIRE

Ashhab Moh’d Sami

2017-01-01

The 12-cyliner camless engine breathing process is modeled with artificial neural networks (ANN’s). The inputs to the net are the intake valve lift (IVL) and intake valve closing timing (IVC) whereas the output of the net is the cylinder air charge (CAC). The ANN is trained with data collected from an engine simulation model which is based on thermodynamics principles and calibrated against real engine data. A method for adapting single-output feed-forward neural networks is proposed and appl...

Critical Neural Substrates for Correcting Unexpected Trajectory Errors and Learning from Them

Science.gov (United States)

Mutha, Pratik K.; Sainburg, Robert L.; Haaland, Kathleen Y.

2011-01-01

Our proficiency at any skill is critically dependent on the ability to monitor our performance, correct errors and adapt subsequent movements so that errors are avoided in the future. In this study, we aimed to dissociate the neural substrates critical for correcting unexpected trajectory errors and learning to adapt future movements based on…

Competition and Cooperation in Neural Nets : U.S.-Japan Joint Seminar

CERN Document Server

Arbib, Michael

1982-01-01

The human brain, wi th its hundred billion or more neurons, is both one of the most complex systems known to man and one of the most important. The last decade has seen an explosion of experimental research on the brain, but little theory of neural networks beyond the study of electrical properties of membranes and small neural circuits. Nonetheless, a number of workers in Japan, the United States and elsewhere have begun to contribute to a theory which provides techniques of mathematical analysis and computer simulation to explore properties of neural systems containing immense numbers of neurons. Recently, it has been gradually recognized that rather independent studies of the dynamics of pattern recognition, pattern format::ion, motor control, self-organization, etc. , in neural systems do in fact make use of common methods. We find that a "competition and cooperation" type of interaction plays a fundamental role in parallel information processing in the brain. The present volume brings together 23 papers ...

Uncovering the neural mechanisms underlying learning from tests.

Directory of Open Access Journals (Sweden)

Xiaonan L Liu

Full Text Available People learn better when re-study opportunities are replaced with tests. While researchers have begun to speculate on why testing is superior to study, few studies have directly examined the neural underpinnings of this effect. In this fMRI study, participants engaged in a study phase to learn arbitrary word pairs, followed by a cued recall test (recall second half of pair when cued with first word of pair, re-study of each pair, and finally another cycle of cued recall tests. Brain activation patterns during the first test (recall of the studied pairs predicts performance on the second test. Importantly, while subsequent memory analyses of encoding trials also predict later accuracy, the brain regions involved in predicting later memory success are more extensive for activity during retrieval (testing than during encoding (study. Those additional regions that predict subsequent memory based on their activation at test but not at encoding may be key to understanding the basis of the testing effect.

Using repetitive transcranial magnetic stimulation to study the underlying neural mechanisms of human motor learning and memory.

Science.gov (United States)

Censor, Nitzan; Cohen, Leonardo G

2011-01-01

In the last two decades, there has been a rapid development in the research of the physiological brain mechanisms underlying human motor learning and memory. While conventional memory research performed on animal models uses intracellular recordings, microfusion of protein inhibitors to specific brain areas and direct induction of focal brain lesions, human research has so far utilized predominantly behavioural approaches and indirect measurements of neural activity. Repetitive transcranial magnetic stimulation (rTMS), a safe non-invasive brain stimulation technique, enables the study of the functional role of specific cortical areas by evaluating the behavioural consequences of selective modulation of activity (excitation or inhibition) on memory generation and consolidation, contributing to the understanding of the neural substrates of motor learning. Depending on the parameters of stimulation, rTMS can also facilitate learning processes, presumably through purposeful modulation of excitability in specific brain regions. rTMS has also been used to gain valuable knowledge regarding the timeline of motor memory formation, from initial encoding to stabilization and long-term retention. In this review, we summarize insights gained using rTMS on the physiological and neural mechanisms of human motor learning and memory. We conclude by suggesting possible future research directions, some with direct clinical implications.

Neural network regulation driven by autonomous neural firings

Science.gov (United States)

Cho, Myoung Won

2016-07-01

Biological neurons naturally fire spontaneously due to the existence of a noisy current. Such autonomous firings may provide a driving force for network formation because synaptic connections can be modified due to neural firings. Here, we study the effect of autonomous firings on network formation. For the temporally asymmetric Hebbian learning, bidirectional connections lose their balance easily and become unidirectional ones. Defining the difference between reciprocal connections as new variables, we could express the learning dynamics as if Ising model spins interact with each other in magnetism. We present a theoretical method to estimate the interaction between the new variables in a neural system. We apply the method to some network systems and find some tendencies of autonomous neural network regulation.

Statistical Discriminability Estimation for Pattern Classification Based on Neural Incremental Attribute Learning

DEFF Research Database (Denmark)

Wang, Ting; Guan, Sheng-Uei; Puthusserypady, Sadasivan

2014-01-01

Feature ordering is a significant data preprocessing method in Incremental Attribute Learning (IAL), a novel machine learning approach which gradually trains features according to a given order. Previous research has shown that, similar to feature selection, feature ordering is also important based...... estimation. Moreover, a criterion that summarizes all the produced values of AD is employed with a GA (Genetic Algorithm)-based approach to obtain the optimum feature ordering for classification problems based on neural networks by means of IAL. Compared with the feature ordering obtained by other approaches...

Modularity and Sparsity: Evolution of Neural Net Controllers in Physically Embodied Robots

Directory of Open Access Journals (Sweden)

Nicholas Livingston

2016-12-01

Full Text Available While modularity is thought to be central for the evolution of complexity and evolvability, it remains unclear how systems boot-strap themselves into modularity from random or fully integrated starting conditions. Clune et al. (2013 suggested that a positive correlation between sparsity and modularity is the prime cause of this transition. We sought to test the generality of this modularity-sparsity hypothesis by testing it for the first time in physically embodied robots. A population of ten Tadros — autonomous, surface-swimming robots propelled by a flapping tail — was used. Individuals varied only in the structure of their neural net control, a 2 x 6 x 2 network with recurrence in the hidden layer. Each of the 60 possible connections was coded in the genome, and could achieve one of three states: -1, 0, 1. Inputs were two light-dependent resistors and outputs were two motor control variables to the flapping tail, one for the frequency of the flapping and the other for the turning offset. Each Tadro was tested separately in a circular tank lit by a single overhead light source. Fitness was the amount of light gathered by a vertically oriented sensor that was disconnected from the controller net. Reproduction was asexual, with the top performer cloned and then all individuals entered into a roulette wheel selection process, with genomes mutated to create the offspring. The starting population of networks was randomly generated. Over ten generations, the population’s mean fitness increased two-fold. This evolution occurred in spite of an unintentional integer overflow problem in recurrent nodes in the hidden layer that caused outputs to oscillate. Our investigation of the oscillatory behavior showed that the mutual information of inputs and outputs was sufficient for the reactive behaviors observed. While we had predicted that both modularity and sparsity would follow the same trend as fitness, neither did so. Instead, selection gradients

Deep Learning at Chest Radiography: Automated Classification of Pulmonary Tuberculosis by Using Convolutional Neural Networks.

Science.gov (United States)

Lakhani, Paras; Sundaram, Baskaran

2017-08-01

Purpose To evaluate the efficacy of deep convolutional neural networks (DCNNs) for detecting tuberculosis (TB) on chest radiographs. Materials and Methods Four deidentified HIPAA-compliant datasets were used in this study that were exempted from review by the institutional review board, which consisted of 1007 posteroanterior chest radiographs. The datasets were split into training (68.0%), validation (17.1%), and test (14.9%). Two different DCNNs, AlexNet and GoogLeNet, were used to classify the images as having manifestations of pulmonary TB or as healthy. Both untrained and pretrained networks on ImageNet were used, and augmentation with multiple preprocessing techniques. Ensembles were performed on the best-performing algorithms. For cases where the classifiers were in disagreement, an independent board-certified cardiothoracic radiologist blindly interpreted the images to evaluate a potential radiologist-augmented workflow. Receiver operating characteristic curves and areas under the curve (AUCs) were used to assess model performance by using the DeLong method for statistical comparison of receiver operating characteristic curves. Results The best-performing classifier had an AUC of 0.99, which was an ensemble of the AlexNet and GoogLeNet DCNNs. The AUCs of the pretrained models were greater than that of the untrained models (P chest radiography with an AUC of 0.99. A radiologist-augmented approach for cases where there was disagreement among the classifiers further improved accuracy. © RSNA, 2017.

Results from a MA16-based neural trigger in an experiment looking for beauty

International Nuclear Information System (INIS)

Baldanza, C.; Beichter, J.; Bisi, F.; Bruels, N.; Bruschini, C.; Cotta-Ramusino, A.; D'Antone, I.; Malferrari, L.; Mazzanti, P.; Musico, P.; Novelli, P.; Odorici, F.; Odorico, R.; Passaseo, M.; Zuffa, M.

1996-01-01

Results from a neural-network trigger based on the digital MA16 chip of Siemens are reported. The neural trigger has been applied to data from the WA92 experiment, looking for beauty particles, which have been collected during a run in which a neural trigger module based on Intel's analog neural chip ETANN operated, as already reported. The MA16 board hosting the chip has a 16-bit I/O precision and a 53-bit precision for internal calculations. It operated at 50 MHz, yielding a response time for a 16 input-variable net of 3 μs for a Fisher discriminant (1-layer net) and of 6 μs for a 2-layer net. Results are compared with those previously obtained with the ETANN trigger. (orig.)

Results from a MA16-based neural trigger in an experiment looking for beauty

Energy Technology Data Exchange (ETDEWEB)

Baldanza, C. [Istituto Nazionale di Fisica Nucleare, Bologna (Italy); Beichter, J. [Siemens AG, ZFE T ME2, 81730 Munich (Germany); Bisi, F. [Istituto Nazionale di Fisica Nucleare, Bologna (Italy); Bruels, N. [Siemens AG, ZFE T ME2, 81730 Munich (Germany); Bruschini, C. [INFN/Genoa, Via Dodecaneso 33, 16146 Genoa (Italy); Cotta-Ramusino, A. [Istituto Nazionale di Fisica Nucleare, Bologna (Italy); D`Antone, I. [Istituto Nazionale di Fisica Nucleare, Bologna (Italy); Malferrari, L. [Istituto Nazionale di Fisica Nucleare, Bologna (Italy); Mazzanti, P. [Istituto Nazionale di Fisica Nucleare, Bologna (Italy); Musico, P. [INFN/Genoa, Via Dodecaneso 33, 16146 Genoa (Italy); Novelli, P. [INFN/Genoa, Via Dodecaneso 33, 16146 Genoa (Italy); Odorici, F. [Istituto Nazionale di Fisica Nucleare, Bologna (Italy); Odorico, R. [Istituto Nazionale di Fisica Nucleare, Bologna (Italy); Passaseo, M. [CERN, 1211 Geneva 23 (Switzerland); Zuffa, M. [Istituto Nazionale di Fisica Nucleare, Bologna (Italy)

1996-07-11

Results from a neural-network trigger based on the digital MA16 chip of Siemens are reported. The neural trigger has been applied to data from the WA92 experiment, looking for beauty particles, which have been collected during a run in which a neural trigger module based on Intel`s analog neural chip ETANN operated, as already reported. The MA16 board hosting the chip has a 16-bit I/O precision and a 53-bit precision for internal calculations. It operated at 50 MHz, yielding a response time for a 16 input-variable net of 3 {mu}s for a Fisher discriminant (1-layer net) and of 6 {mu}s for a 2-layer net. Results are compared with those previously obtained with the ETANN trigger. (orig.).

Transfer Learning with Convolutional Neural Networks for SAR Ship Recognition

Science.gov (United States)

Zhang, Di; Liu, Jia; Heng, Wang; Ren, Kaijun; Song, Junqiang

2018-03-01

Ship recognition is the backbone of marine surveillance systems. Recent deep learning methods, e.g. Convolutional Neural Networks (CNNs), have shown high performance for optical images. Learning CNNs, however, requires a number of annotated samples to estimate numerous model parameters, which prevents its application to Synthetic Aperture Radar (SAR) images due to the limited annotated training samples. Transfer learning has been a promising technique for applications with limited data. To this end, a novel SAR ship recognition method based on CNNs with transfer learning has been developed. In this work, we firstly start with a CNNs model that has been trained in advance on Moving and Stationary Target Acquisition and Recognition (MSTAR) database. Next, based on the knowledge gained from this image recognition task, we fine-tune the CNNs on a new task to recognize three types of ships in the OpenSARShip database. The experimental results show that our proposed approach can obviously increase the recognition rate comparing with the result of merely applying CNNs. In addition, compared to existing methods, the proposed method proves to be very competitive and can learn discriminative features directly from training data instead of requiring pre-specification or pre-selection manually.

Differential neural substrates of working memory and cognitive skill learning in healthy young volunteers

International Nuclear Information System (INIS)

Cho, Sang Soo; Lee, Eun Ju; Yoon, Eun Jin; Kim, Yu Kyeong; Lee, Won Woo; Kim, Sang Eun

2005-01-01

It is known that different neural circuits are involved in working memory and cognitive skill learning that represent explicit and implicit memory functions, respectively. In the present study, we investigated the metabolic correlates of working memory and cognitive skill learning with correlation analysis of FDG PET images. Fourteen right-handed healthy subjects (age, 24 ± 2 yr; 5 males and 9 females) underwent brain FDG PET and neuropsychological testing. Two-back task and weather prediction task were used for the evaluation of working memory and cognitive skill learning, respectively, Correlation between regional glucose metabolism and cognitive task performance was examined using SPM99. A significant positive correlation between 2-back task performance and regional glucose metabolism was found in the prefrontal regions and superior temporal gyri bilaterally. In the first term of weather prediction task the task performance correlated positively with glucose metabolism in the bilateral prefrontal areas, left middle temporal and posterior cingulate gyri, and left thalamus. In the second and third terms of the task, the correlation found in the prefrontal areas, superior temporal and anterior cingulate gyri bilaterally, right insula, left parahippocampal gyrus, and right caudate nucleus. We identified the neural substrates that are related with performance of working memory and cognitive skill learning. These results indicate that brain regions associated with the explicit memory system are recruited in early periods of cognitive skill learning, but additional brain regions including caudate nucleus are involved in late periods of cognitive skill learning

Differential neural substrates of working memory and cognitive skill learning in healthy young volunteers

Energy Technology Data Exchange (ETDEWEB)

Cho, Sang Soo; Lee, Eun Ju; Yoon, Eun Jin; Kim, Yu Kyeong; Lee, Won Woo; Kim, Sang Eun [Seoul National Univ. College of Medicine, Seoul (Korea, Republic of)

2005-07-01

It is known that different neural circuits are involved in working memory and cognitive skill learning that represent explicit and implicit memory functions, respectively. In the present study, we investigated the metabolic correlates of working memory and cognitive skill learning with correlation analysis of FDG PET images. Fourteen right-handed healthy subjects (age, 24 {+-} 2 yr; 5 males and 9 females) underwent brain FDG PET and neuropsychological testing. Two-back task and weather prediction task were used for the evaluation of working memory and cognitive skill learning, respectively, Correlation between regional glucose metabolism and cognitive task performance was examined using SPM99. A significant positive correlation between 2-back task performance and regional glucose metabolism was found in the prefrontal regions and superior temporal gyri bilaterally. In the first term of weather prediction task the task performance correlated positively with glucose metabolism in the bilateral prefrontal areas, left middle temporal and posterior cingulate gyri, and left thalamus. In the second and third terms of the task, the correlation found in the prefrontal areas, superior temporal and anterior cingulate gyri bilaterally, right insula, left parahippocampal gyrus, and right caudate nucleus. We identified the neural substrates that are related with performance of working memory and cognitive skill learning. These results indicate that brain regions associated with the explicit memory system are recruited in early periods of cognitive skill learning, but additional brain regions including caudate nucleus are involved in late periods of cognitive skill learning.

Exploration Of Deep Learning Algorithms Using Openacc Parallel Programming Model

KAUST Repository

Hamam, Alwaleed A.

2017-03-13

Deep learning is based on a set of algorithms that attempt to model high level abstractions in data. Specifically, RBM is a deep learning algorithm that used in the project to increase it\\'s time performance using some efficient parallel implementation by OpenACC tool with best possible optimizations on RBM to harness the massively parallel power of NVIDIA GPUs. GPUs development in the last few years has contributed to growing the concept of deep learning. OpenACC is a directive based ap-proach for computing where directives provide compiler hints to accelerate code. The traditional Restricted Boltzmann Ma-chine is a stochastic neural network that essentially perform a binary version of factor analysis. RBM is a useful neural net-work basis for larger modern deep learning model, such as Deep Belief Network. RBM parameters are estimated using an efficient training method that called Contrastive Divergence. Parallel implementation of RBM is available using different models such as OpenMP, and CUDA. But this project has been the first attempt to apply OpenACC model on RBM.

Exploration Of Deep Learning Algorithms Using Openacc Parallel Programming Model

KAUST Repository

Hamam, Alwaleed A.; Khan, Ayaz H.

2017-01-01

Deep learning is based on a set of algorithms that attempt to model high level abstractions in data. Specifically, RBM is a deep learning algorithm that used in the project to increase it's time performance using some efficient parallel implementation by OpenACC tool with best possible optimizations on RBM to harness the massively parallel power of NVIDIA GPUs. GPUs development in the last few years has contributed to growing the concept of deep learning. OpenACC is a directive based ap-proach for computing where directives provide compiler hints to accelerate code. The traditional Restricted Boltzmann Ma-chine is a stochastic neural network that essentially perform a binary version of factor analysis. RBM is a useful neural net-work basis for larger modern deep learning model, such as Deep Belief Network. RBM parameters are estimated using an efficient training method that called Contrastive Divergence. Parallel implementation of RBM is available using different models such as OpenMP, and CUDA. But this project has been the first attempt to apply OpenACC model on RBM.

Deciphering the Cognitive and Neural Mechanisms Underlying ...

International Development Research Centre (IDRC) Digital Library (Canada)

Deciphering the Cognitive and Neural Mechanisms Underlying Auditory Learning. This project seeks to understand the brain mechanisms necessary for people to learn to perceive sounds. Neural circuits and learning. The research team will test people with and without musical training to evaluate their capacity to learn ...

Kernel Temporal Differences for Neural Decoding

Science.gov (United States)

Bae, Jihye; Sanchez Giraldo, Luis G.; Pohlmeyer, Eric A.; Francis, Joseph T.; Sanchez, Justin C.; Príncipe, José C.

2015-01-01

We study the feasibility and capability of the kernel temporal difference (KTD)(λ) algorithm for neural decoding. KTD(λ) is an online, kernel-based learning algorithm, which has been introduced to estimate value functions in reinforcement learning. This algorithm combines kernel-based representations with the temporal difference approach to learning. One of our key observations is that by using strictly positive definite kernels, algorithm's convergence can be guaranteed for policy evaluation. The algorithm's nonlinear functional approximation capabilities are shown in both simulations of policy evaluation and neural decoding problems (policy improvement). KTD can handle high-dimensional neural states containing spatial-temporal information at a reasonable computational complexity allowing real-time applications. When the algorithm seeks a proper mapping between a monkey's neural states and desired positions of a computer cursor or a robot arm, in both open-loop and closed-loop experiments, it can effectively learn the neural state to action mapping. Finally, a visualization of the coadaptation process between the decoder and the subject shows the algorithm's capabilities in reinforcement learning brain machine interfaces. PMID:25866504

Neural principles of memory and a neural theory of analogical insight

Science.gov (United States)

Lawson, David I.; Lawson, Anton E.

1993-12-01

Grossberg's principles of neural modeling are reviewed and extended to provide a neural level theory to explain how analogies greatly increase the rate of learning and can, in fact, make learning and retention possible. In terms of memory, the key point is that the mind is able to recognize and recall when it is able to match sensory input from new objects, events, or situations with past memory records of similar objects, events, or situations. When a match occurs, an adaptive resonance is set up in which the synaptic strengths of neurons are increased; thus a long term record of the new input is formed in memory. Systems of neurons called outstars and instars are presumably the underlying units that enable this to occur. Analogies can greatly facilitate learning and retention because they activate the outstars (i.e., the cells that are sampling the to-be-learned pattern) and cause the neural activity to grow exponentially by forming feedback loops. This increased activity insures the boost in synaptic strengths of neurons, thus causing storage and retention in long-term memory (i.e., learning).

STUDY ON THE CLASSIFICATION OF GAOFEN-3 POLARIMETRIC SAR IMAGES USING DEEP NEURAL NETWORK

Directory of Open Access Journals (Sweden)

J. Zhang

2018-04-01

Full Text Available Polarimetric Synthetic Aperture Radar（POLSAR） imaging principle determines that the image quality will be affected by speckle noise. So the recognition accuracy of traditional image classification methods will be reduced by the effect of this interference. Since the date of submission, Deep Convolutional Neural Network impacts on the traditional image processing methods and brings the field of computer vision to a new stage with the advantages of a strong ability to learn deep features and excellent ability to fit large datasets. Based on the basic characteristics of polarimetric SAR images, the paper studied the types of the surface cover by using the method of Deep Learning. We used the fully polarimetric SAR features of different scales to fuse RGB images to the GoogLeNet model based on convolution neural network Iterative training, and then use the trained model to test the classification of data validation.First of all, referring to the optical image, we mark the surface coverage type of GF-3 POLSAR image with 8m resolution, and then collect the samples according to different categories. To meet the GoogLeNet model requirements of 256 × 256 pixel image input and taking into account the lack of full-resolution SAR resolution, the original image should be pre-processed in the process of resampling. In this paper, POLSAR image slice samples of different scales with sampling intervals of 2 m and 1 m to be trained separately and validated by the verification dataset. Among them, the training accuracy of GoogLeNet model trained with resampled 2-m polarimetric SAR image is 94.89 %, and that of the trained SAR image with resampled 1 m is 92.65 %.

Study on the Classification of GAOFEN-3 Polarimetric SAR Images Using Deep Neural Network

Science.gov (United States)

Zhang, J.; Zhang, J.; Zhao, Z.

2018-04-01

Polarimetric Synthetic Aperture Radar (POLSAR) imaging principle determines that the image quality will be affected by speckle noise. So the recognition accuracy of traditional image classification methods will be reduced by the effect of this interference. Since the date of submission, Deep Convolutional Neural Network impacts on the traditional image processing methods and brings the field of computer vision to a new stage with the advantages of a strong ability to learn deep features and excellent ability to fit large datasets. Based on the basic characteristics of polarimetric SAR images, the paper studied the types of the surface cover by using the method of Deep Learning. We used the fully polarimetric SAR features of different scales to fuse RGB images to the GoogLeNet model based on convolution neural network Iterative training, and then use the trained model to test the classification of data validation.First of all, referring to the optical image, we mark the surface coverage type of GF-3 POLSAR image with 8m resolution, and then collect the samples according to different categories. To meet the GoogLeNet model requirements of 256 × 256 pixel image input and taking into account the lack of full-resolution SAR resolution, the original image should be pre-processed in the process of resampling. In this paper, POLSAR image slice samples of different scales with sampling intervals of 2 m and 1 m to be trained separately and validated by the verification dataset. Among them, the training accuracy of GoogLeNet model trained with resampled 2-m polarimetric SAR image is 94.89 %, and that of the trained SAR image with resampled 1 m is 92.65 %.

On open questions in the geometric approach to structural learning Bayesian nets

Czech Academy of Sciences Publication Activity Database

Studený, Milan; Vomlel, Jiří

2011-01-01

Roč. 52, č. 5 (2011), s. 627-640 ISSN 0888-613X. [Workshop on Uncertainty Processing WUPES'09 /8./. Liblice, 19.09.2009-23.09.2009] R&D Projects: GA MŠk(CZ) 1M0572; GA ČR GA201/08/0539; GA ČR GEICC/08/E010 Grant - others:GA MŠk(CZ) 2C06019 Institutional research plan: CEZ:AV0Z10750506 Keywords : structural learning Bayesian nets * standard imset * polytope * geometric neighborhood * differential imset Subject RIV: BA - General Mathematics Impact factor: 1.948, year: 2011 http://library.utia.cas.cz/separaty/2011/MTR/studeny-0358907. pdf