ARIMA-Based Time Series Model of Stochastic Wind Power Generation
Chen, Peiyuan; Pedersen, Troels; Bak-Jensen, Birgitte;
2010-01-01
This paper proposes a stochastic wind power model based on an autoregressive integrated moving average (ARIMA) process. The model takes into account the nonstationarity and physical limits of stochastic wind power generation. The model is constructed based on wind power measurement of one year from...... the Nysted offshore wind farm in Denmark. The proposed limited-ARIMA (LARIMA) model introduces a limiter and characterizes the stochastic wind power generation by mean level, temporal correlation and driving noise. The model is validated against the measurement in terms of temporal correlation and...
Forecasting inflation in Montenegro using univariate time series models
Milena Lipovina-Božović
2015-04-01
Full Text Available The analysis of price trends and their prognosis is one of the key tasks of the economic authorities in each country. Due to the nature of the Montenegrin economy as small and open economy with euro as currency, forecasting inflation is very specific which is more difficult due to low quality of the data. This paper analyzes the utility and applicability of univariate time series models for forecasting price index in Montenegro. Data analysis of key macroeconomic movements in previous decades indicates the presence of many possible determinants that could influence forecasting result. This paper concludes that the forecasting models (ARIMA based only on its own previous values cannot adequately cover the key factors that determine the price level in the future, probably because of the existence of numerous external factors that influence the price movement in Montenegro.
Moskowitz, Tobias J.; Ooi, Yao Hua; Heje Pedersen, Lasse
2012-01-01
We document significant “time series momentum” in equity index, currency, commodity, and bond futures for each of the 58 liquid instruments we consider. We find persistence in returns for one to 12 months that partially reverses over longer horizons, consistent with sentiment theories of initial...... under-reaction and delayed over-reaction. A diversified portfolio of time series momentum strategies across all asset classes delivers substantial abnormal returns with little exposure to standard asset pricing factors and performs best during extreme markets. Examining the trading activities...... of speculators and hedgers, we find that speculators profit from time series momentum at the expense of hedgers....
Loredo, Thomas
The key, central objectives of the proposed Time Series Explorer project are to develop an organized collection of software tools for analysis of time series data in current and future NASA astrophysics data archives, and to make the tools available in two ways: as a library (the Time Series Toolbox) that individual science users can use to write their own data analysis pipelines, and as an application (the Time Series Automaton) providing an accessible, data-ready interface to many Toolbox algorithms, facilitating rapid exploration and automatic processing of time series databases. A number of time series analysis methods will be implemented, including techniques that range from standard ones to state-of-the-art developments by the proposers and others. Most of the algorithms will be able to handle time series data subject to real-world problems such as data gaps, sampling that is otherwise irregular, asynchronous sampling (in multi-wavelength settings), and data with non-Gaussian measurement errors. The proposed research responds to the ADAP element supporting the development of tools for mining the vast reservoir of information residing in NASA databases. The tools that will be provided to the community of astronomers studying variability of astronomical objects (from nearby stars and extrasolar planets, through galactic and extragalactic sources) will revolutionize the quality of timing analyses that can be carried out, and greatly enhance the scientific throughput of all NASA astrophysics missions past, present, and future. The Automaton will let scientists explore time series - individual records or large data bases -- with the most informative and useful analysis methods available, without having to develop the tools themselves or understand the computational details. Both elements, the Toolbox and the Automaton, will enable deep but efficient exploratory time series data analysis, which is why we have named the project the Time Series Explorer. Science
Multivariate Time Series Search
National Aeronautics and Space Administration — Multivariate Time-Series (MTS) are ubiquitous, and are generated in areas as disparate as sensor recordings in aerospace systems, music and video streams, medical...
Hisdal, H.; Holmqvist, E.; Hyvärinen, V.; Jónsson, P.; Kuusisto, E.; Larsen, S. E.; Lindström, G.; Ovesen, N. B.; Roald, L. A.
Awareness that emission of greenhouse gases will raise the global temperature and change the climate has led to studies trying to identify such changes in long-term climate and hydrologic time series. This report, written by the......Awareness that emission of greenhouse gases will raise the global temperature and change the climate has led to studies trying to identify such changes in long-term climate and hydrologic time series. This report, written by the...
Fischer, Paul; Hilbert, Astrid
2012-01-01
We introduce a platform which supplies an easy-to-handle, interactive, extendable, and fast analysis tool for time series analysis. In contrast to other software suits like Maple, Matlab, or R, which use a command-line-like interface and where the user has to memorize/look-up the appropriate...... commands, our application is select-and-click-driven. It allows to derive many different sequences of deviations for a given time series and to visualize them in different ways in order to judge their expressive power and to reuse the procedure found. For many transformations or model-ts, the user may...... choose between manual and automated parameter selection. The user can dene new transformations and add them to the system. The application contains efficient implementations of advanced and recent techniques for time series analysis including techniques related to extreme value analysis and filtering...
Madsen, Henrik
2007-01-01
""In this book the author gives a detailed account of estimation, identification methodologies for univariate and multivariate stationary time-series models. The interesting aspect of this introductory book is that it contains several real data sets and the author made an effort to explain and motivate the methodology with real data. … this introductory book will be interesting and useful not only to undergraduate students in the UK universities but also to statisticians who are keen to learn time-series techniques and keen to apply them. I have no hesitation in recommending the book.""-Journa
Woodward, Wayne A; Elliott, Alan C
2011-01-01
""There is scarcely a standard technique that the reader will find left out … this book is highly recommended for those requiring a ready introduction to applicable methods in time series and serves as a useful resource for pedagogical purposes.""-International Statistical Review (2014), 82""Current time series theory for practice is well summarized in this book.""-Emmanuel Parzen, Texas A&M University""What an extraordinary range of topics covered, all very insightfully. I like [the authors'] innovations very much, such as the AR factor table.""-David Findley, U.S. Census Bureau (retired)""…
Liang, X San
2014-01-01
Given two time series, can one tell, in a rigorous and quantitative way, the cause and effect between them? Based on a recently rigorized physical notion namely information flow, we arrive at a concise formula and give this challenging question, which is of wide concern in different disciplines, a positive answer. Here causality is measured by the time rate of change of information flowing from one series, say, X2, to another, X1. The measure is asymmetric between the two parties and, particularly, if the process underlying X1 does not depend on X2, then the resulting causality from X2 to X1 vanishes. The formula is tight in form, involving only the commonly used statistics, sample covariances. It has been validated with touchstone series purportedly generated with one-way causality. It has also been applied to the investigation of real world problems; an example presented here is the cause-effect relation between two climate modes, El Ni\\~no and Indian Ocean Dipole, which have been linked to the hazards in f...
Many human-related activities show power-law decaying interevent time distribution with exponents usually varying between 1 and 2. We study a simple task-queuing model, which produces bursty time series due to the non-trivial dynamics of the task list. The model is characterized by a priority distribution as an input parameter, which describes the choice procedure from the list. We give exact results on the asymptotic behaviour of the model and we show that the interevent time distribution is power-law decaying for any kind of input distributions that remain normalizable in the infinite list limit, with exponents tunable between 1 and 2. The model satisfies a scaling law between the exponents of interevent time distribution (β) and autocorrelation function (α): α + β = 2. This law is general for renewal processes with power-law decaying interevent time distribution. We conclude that slowly decaying autocorrelation function indicates long-range dependence only if the scaling law is violated. (paper)
GPS Position Time Series @ JPL
Owen, Susan; Moore, Angelyn; Kedar, Sharon; Liu, Zhen; Webb, Frank; Heflin, Mike; Desai, Shailen
2013-01-01
Different flavors of GPS time series analysis at JPL - Use same GPS Precise Point Positioning Analysis raw time series - Variations in time series analysis/post-processing driven by different users. center dot JPL Global Time Series/Velocities - researchers studying reference frame, combining with VLBI/SLR/DORIS center dot JPL/SOPAC Combined Time Series/Velocities - crustal deformation for tectonic, volcanic, ground water studies center dot ARIA Time Series/Coseismic Data Products - Hazard monitoring and response focused center dot ARIA data system designed to integrate GPS and InSAR - GPS tropospheric delay used for correcting InSAR - Caltech's GIANT time series analysis uses GPS to correct orbital errors in InSAR - Zhen Liu's talking tomorrow on InSAR Time Series analysis
Bootstrapping High Dimensional Time Series
Zhang, Xianyang; Cheng, Guang
2014-01-01
This article studies bootstrap inference for high dimensional weakly dependent time series in a general framework of approximately linear statistics. The following high dimensional applications are covered: (1) uniform confidence band for mean vector; (2) specification testing on the second order property of time series such as white noise testing and bandedness testing of covariance matrix; (3) specification testing on the spectral property of time series. In theory, we first derive a Gaussi...
Autoencoding Time Series for Visualisation
Gianniotis, Nikolaos; Kügler, Dennis; Tino, Peter; Polsterer, Kai; Misra, Ranjeev
2015-01-01
We present an algorithm for the visualisation of time series. To that end we employ echo state networks to convert time series into a suitable vector representation which is capable of capturing the latent dynamics of the time series. Subsequently, the obtained vector representations are put through an autoencoder and the visualisation is constructed using the activations of the bottleneck. The crux of the work lies with defining an objective function that quantifies the reconstruction error ...
Allan, Alasdair
2014-06-01
FROG performs time series analysis and display. It provides a simple user interface for astronomers wanting to do time-domain astrophysics but still offers the powerful features found in packages such as PERIOD (ascl:1406.005). FROG includes a number of tools for manipulation of time series. Among other things, the user can combine individual time series, detrend series (multiple methods) and perform basic arithmetic functions. The data can also be exported directly into the TOPCAT (ascl:1101.010) application for further manipulation if needed.
Advances in time series forecasting
Cagdas, Hakan Aladag
2012-01-01
Readers will learn how these methods work and how these approaches can be used to forecast real life time series. The hybrid forecasting model is also explained. Data presented in this e-book is problem based and is taken from real life situations. It is a valuable resource for students, statisticians and working professionals interested in advanced time series analysis.
Models for dependent time series
Tunnicliffe Wilson, Granville; Haywood, John
2015-01-01
Models for Dependent Time Series addresses the issues that arise and the methodology that can be applied when the dependence between time series is described and modeled. Whether you work in the economic, physical, or life sciences, the book shows you how to draw meaningful, applicable, and statistically valid conclusions from multivariate (or vector) time series data.The first four chapters discuss the two main pillars of the subject that have been developed over the last 60 years: vector autoregressive modeling and multivariate spectral analysis. These chapters provide the foundational mater
Autoencoding Time Series for Visualisation
Gianniotis, Nikolaos; Tino, Peter; Polsterer, Kai; Misra, Ranjeev
2015-01-01
We present an algorithm for the visualisation of time series. To that end we employ echo state networks to convert time series into a suitable vector representation which is capable of capturing the latent dynamics of the time series. Subsequently, the obtained vector representations are put through an autoencoder and the visualisation is constructed using the activations of the bottleneck. The crux of the work lies with defining an objective function that quantifies the reconstruction error of these representations in a principled manner. We demonstrate the method on synthetic and real data.
Time Series with Tailored Nonlinearities
Raeth, C
2015-01-01
It is demonstrated how to generate time series with tailored nonlinearities by inducing well- defined constraints on the Fourier phases. Correlations between the phase information of adjacent phases and (static and dynamic) measures of nonlinearities are established and their origin is explained. By applying a set of simple constraints on the phases of an originally linear and uncor- related Gaussian time series, the observed scaling behavior of the intensity distribution of empirical time series can be reproduced. The power law character of the intensity distributions being typical for e.g. turbulence and financial data can thus be explained in terms of phase correlations.
Multivariate Time Series Similarity Searching
Jimin Wang; Yuelong Zhu; Shijin Li; Dingsheng Wan; Pengcheng Zhang
2014-01-01
Multivariate time series (MTS) datasets are very common in various financial, multimedia, and hydrological fields. In this paper, a dimension-combination method is proposed to search similar sequences for MTS. Firstly, the similarity of single-dimension series is calculated; then the overall similarity of the MTS is obtained by synthesizing each of the single-dimension similarity based on weighted BORDA voting method. The dimension-combination method could use the existing similarity searchin...
Random time series in astronomy.
Vaughan, Simon
2013-02-13
Progress in astronomy comes from interpreting the signals encoded in the light received from distant objects: the distribution of light over the sky (images), over photon wavelength (spectrum), over polarization angle and over time (usually called light curves by astronomers). In the time domain, we see transient events such as supernovae, gamma-ray bursts and other powerful explosions; we see periodic phenomena such as the orbits of planets around nearby stars, radio pulsars and pulsations of stars in nearby galaxies; and we see persistent aperiodic variations ('noise') from powerful systems such as accreting black holes. I review just a few of the recent and future challenges in the burgeoning area of time domain astrophysics, with particular attention to persistently variable sources, the recovery of reliable noise power spectra from sparsely sampled time series, higher order properties of accreting black holes, and time delays and correlations in multi-variate time series. PMID:23277606
Stochastic Time-Series Spectroscopy
Scoville, John
2015-01-01
Spectroscopically measuring low levels of non-equilibrium phenomena (e.g. emission in the presence of a large thermal background) can be problematic due to an unfavorable signal-to-noise ratio. An approach is presented to use time-series spectroscopy to separate non-equilibrium quantities from slowly varying equilibria. A stochastic process associated with the non-equilibrium part of the spectrum is characterized in terms of its central moments or cumulants, which may vary over time. This parameterization encodes information about the non-equilibrium behavior of the system. Stochastic time-series spectroscopy (STSS) can be implemented at very little expense in many settings since a series of scans are typically recorded in order to generate a low-noise averaged spectrum. Higher moments or cumulants may be readily calculated from this series, enabling the observation of quantities that would be difficult or impossible to determine from an average spectrum or from prinicipal components analysis (PCA). This meth...
ACCURATE TIME SERIES CLASSIFICATION USING SHAPELETS
M. Arathi; A. GOVARDHAN
2014-01-01
Time series data are sequences of values measured o ver time. One of the most recent approaches to classification of time series data is to find shape lets within a data set. Time series shapelets are time series subsequences which represent a class. In order to compare two time series sequences, existing work use s Euclidean distance measure. The problem with Euclid ean distance is that it requires data to be standardized if scales ...
Nonlinear time series analysis methods and applications
Diks, Cees
1999-01-01
Methods of nonlinear time series analysis are discussed from a dynamical systems perspective on the one hand, and from a statistical perspective on the other. After giving an informal overview of the theory of dynamical systems relevant to the analysis of deterministic time series, time series generated by nonlinear stochastic systems and spatio-temporal dynamical systems are considered. Several statistical methods for the analysis of nonlinear time series are presented and illustrated with applications to physical and physiological time series.
Trend prediction of chaotic time series
Li Aiguo; Zhao Cai; Li Zhanhuai
2007-01-01
To predict the trend of chaotic time series in time series analysis and time series data mining fields, a novel predicting algorithm of chaotic time series trend is presented, and an on-line segmenting algorithm is proposed to convert a time series into a binary string according to ascending or descending trend of each subsequence. The on-line segmenting algorithm is independent of the prior knowledge about time series. The naive Bayesian algorithm is then employed to predict the trend of chaotic time series according to the binary string. The experimental results of three chaotic time series demonstrate that the proposed method predicts the ascending or descending trend of chaotic time series with few error.
A Course in Time Series Analysis
Peña, Daniel; Tsay, Ruey S
2011-01-01
New statistical methods and future directions of research in time series A Course in Time Series Analysis demonstrates how to build time series models for univariate and multivariate time series data. It brings together material previously available only in the professional literature and presents a unified view of the most advanced procedures available for time series model building. The authors begin with basic concepts in univariate time series, providing an up-to-date presentation of ARIMA models, including the Kalman filter, outlier analysis, automatic methods for building ARIMA models, a
Effective Feature Preprocessing for Time Series Forecasting
Zhao, Junhua; Dong, Zhaoyang; Xu, Zhao
2006-01-01
Time series forecasting is an important area in data mining research. Feature preprocessing techniques have significant influence on forecasting accuracy, therefore are essential in a forecasting model. Although several feature preprocessing techniques have been applied in time series forecasting...
Effective Feature Preprocessing for Time Series Forecasting
Zhao, Junhua; Dong, Zhaoyang; Xu, Zhao
Time series forecasting is an important area in data mining research. Feature preprocessing techniques have significant influence on forecasting accuracy, therefore are essential in a forecasting model. Although several feature preprocessing techniques have been applied in time series forecasting...
Time Series Analysis and Forecasting by Example
Bisgaard, Soren
2011-01-01
An intuition-based approach enables you to master time series analysis with ease Time Series Analysis and Forecasting by Example provides the fundamental techniques in time series analysis using various examples. By introducing necessary theory through examples that showcase the discussed topics, the authors successfully help readers develop an intuitive understanding of seemingly complicated time series models and their implications. The book presents methodologies for time series analysis in a simplified, example-based approach. Using graphics, the authors discuss each presented example in
A review of subsequence time series clustering.
Zolhavarieh, Seyedjamal; Aghabozorgi, Saeed; Teh, Ying Wah
2014-01-01
Clustering of subsequence time series remains an open issue in time series clustering. Subsequence time series clustering is used in different fields, such as e-commerce, outlier detection, speech recognition, biological systems, DNA recognition, and text mining. One of the useful fields in the domain of subsequence time series clustering is pattern recognition. To improve this field, a sequence of time series data is used. This paper reviews some definitions and backgrounds related to subsequence time series clustering. The categorization of the literature reviews is divided into three groups: preproof, interproof, and postproof period. Moreover, various state-of-the-art approaches in performing subsequence time series clustering are discussed under each of the following categories. The strengths and weaknesses of the employed methods are evaluated as potential issues for future studies. PMID:25140332
Data mining in time series databases
Kandel, Abraham; Bunke, Horst
2004-01-01
Adding the time dimension to real-world databases produces Time SeriesDatabases (TSDB) and introduces new aspects and difficulties to datamining and knowledge discovery. This book covers the state-of-the-artmethodology for mining time series databases. The novel data miningmethods presented in the book include techniques for efficientsegmentation, indexing, and classification of noisy and dynamic timeseries. A graph-based method for anomaly detection in time series isdescribed and the book also studies the implications of a novel andpotentially useful representation of time series as strings. Theproblem of detecting changes in data mining models that are inducedfrom temporal databases is additionally discussed.
International Work-Conference on Time Series
Pomares, Héctor
2016-01-01
This volume presents selected peer-reviewed contributions from The International Work-Conference on Time Series, ITISE 2015, held in Granada, Spain, July 1-3, 2015. It discusses topics in time series analysis and forecasting, advanced methods and online learning in time series, high-dimensional and complex/big data time series as well as forecasting in real problems. The International Work-Conferences on Time Series (ITISE) provide a forum for scientists, engineers, educators and students to discuss the latest ideas and implementations in the foundations, theory, models and applications in the field of time series analysis and forecasting. It focuses on interdisciplinary and multidisciplinary research encompassing the disciplines of computer science, mathematics, statistics and econometrics.
Coupling between time series: a network view
Mehraban, Saeed; Zamani, Maryam; Jafari, Gholamreza
2013-01-01
Recently, the visibility graph has been introduced as a novel view for analyzing time series, which maps it to a complex network. In this paper, we introduce new algorithm of visibility, "cross-visibility", which reveals the conjugation of two coupled time series. The correspondence between the two time series is mapped to a network, "the cross-visibility graph", to demonstrate the correlation between them. We applied the algorithm to several correlated and uncorrelated time series, generated by the linear stationary ARFIMA process. The results demonstrate that the cross-visibility graph associated with correlated time series with power-law auto-correlation is scale-free. If the time series are uncorrelated, the degree distribution of their cross-visibility network deviates from power-law. For more clarifying the process, we applied the algorithm to real-world data from the financial trades of two companies, and observed significant small-scale coupling in their dynamics.
Random time series in Astronomy
Vaughan, Simon
2013-01-01
Progress in astronomy comes from interpreting the signals encoded in the light received from distant objects: the distribution of light over the sky (images), over photon wavelength (spectrum), over polarization angle, and over time (usually called light curves by astronomers). In the time domain we see transient events such as supernovae, gamma-ray bursts, and other powerful explosions; we see periodic phenomena such as the orbits of planets around nearby stars, radio pulsars, and pulsations...
Hurst Exponent Analysis of Financial Time Series
无
2001-01-01
Statistical properties of stock market time series and the implication of their Hurst exponents are discussed. Hurst exponents of DJ1A (Dow Jones Industrial Average) components are tested using re-scaled range analysis. In addition to the original stock return series, the linear prediction errors of the daily returns are also tested. Numerical results show that the Hurst exponent analysis can provide some information about the statistical properties of the financial time series.
The Foundations of Modern Time Series Analysis
Mills, Professor Terence C
2011-01-01
This book develops the analysis of Time Series from its formal beginnings in the 1890s through to the publication of Box and Jenkins' watershed publication in 1970, showing how these methods laid the foundations for the modern techniques of Time Series analysis that are in use today.
Lag space estimation in time series modelling
Goutte, Cyril
1997-01-01
The purpose of this article is to investigate some techniques for finding the relevant lag-space, i.e. input information, for time series modelling. This is an important aspect of time series modelling, as it conditions the design of the model through the regressor vector a.k.a. the input layer...
Network structure of multivariate time series
Lacasa, Lucas; Nicosia, Vincenzo; Latora, Vito
2015-10-01
Our understanding of a variety of phenomena in physics, biology and economics crucially depends on the analysis of multivariate time series. While a wide range tools and techniques for time series analysis already exist, the increasing availability of massive data structures calls for new approaches for multidimensional signal processing. We present here a non-parametric method to analyse multivariate time series, based on the mapping of a multidimensional time series into a multilayer network, which allows to extract information on a high dimensional dynamical system through the analysis of the structure of the associated multiplex network. The method is simple to implement, general, scalable, does not require ad hoc phase space partitioning, and is thus suitable for the analysis of large, heterogeneous and non-stationary time series. We show that simple structural descriptors of the associated multiplex networks allow to extract and quantify nontrivial properties of coupled chaotic maps, including the transition between different dynamical phases and the onset of various types of synchronization. As a concrete example we then study financial time series, showing that a multiplex network analysis can efficiently discriminate crises from periods of financial stability, where standard methods based on time-series symbolization often fail.
Homogenising time series: beliefs, dogmas and facts
Domonkos, P.
2011-06-01
In the recent decades various homogenisation methods have been developed, but the real effects of their application on time series are still not known sufficiently. The ongoing COST action HOME (COST ES0601) is devoted to reveal the real impacts of homogenisation methods more detailed and with higher confidence than earlier. As a part of the COST activity, a benchmark dataset was built whose characteristics approach well the characteristics of real networks of observed time series. This dataset offers much better opportunity than ever before to test the wide variety of homogenisation methods, and analyse the real effects of selected theoretical recommendations. Empirical results show that real observed time series usually include several inhomogeneities of different sizes. Small inhomogeneities often have similar statistical characteristics than natural changes caused by climatic variability, thus the pure application of the classic theory that change-points of observed time series can be found and corrected one-by-one is impossible. However, after homogenisation the linear trends, seasonal changes and long-term fluctuations of time series are usually much closer to the reality than in raw time series. Some problems around detecting multiple structures of inhomogeneities, as well as that of time series comparisons within homogenisation procedures are discussed briefly in the study.
Testing Mean Stability of Heteroskedastic Time Series
Violetta Dalla; Liudas Giraitis; Phillips, Peter C. B.
2015-01-01
Time series models are often fitted to the data without preliminary checks for stability of the mean and variance, conditions that may not hold in much economic and financial data, particularly over long periods. Ignoring such shifts may result in fitting models with spurious dynamics that lead to unsupported and controversial conclusions about time dependence, causality, and the effects of unanticipated shocks. In spite of what may seem as obvious differences between a time series of indepen...
Testing mean stability of heteroskedastic time series
Dalla, Violetta; Giraitis, Liudas; Phillips, Peter C. B.
2015-01-01
Time series models are often fitted to the data without preliminary checks for stability of the mean and variance, conditions that may not hold in much economic and financial data, particularly over long periods. Ignoring such shifts may result in fitting models with spurious dynamics that lead to unsupported and controversial conclusions about time dependence, causality, and the effects of unanticipated shocks. In spite of what may seem as obvious differences between a time series of indepen...
Time Series Analysis Using Composite Multiscale Entropy
Kung-Yen Lee; Chun-Chieh Wang; Shiou-Gwo Lin; Chiu-Wen Wu; Shuen-De Wu
2013-01-01
Multiscale entropy (MSE) was recently developed to evaluate the complexity of time series over different time scales. Although the MSE algorithm has been successfully applied in a number of different fields, it encounters a problem in that the statistical reliability of the sample entropy (SampEn) of a coarse-grained series is reduced as a time scale factor is increased. Therefore, in this paper, the concept of a composite multiscale entropy (CMSE) is introduced to overcome this difficulty. S...
Time series modeling, computation, and inference
Prado, Raquel
2010-01-01
The authors systematically develop a state-of-the-art analysis and modeling of time series. … this book is well organized and well written. The authors present various statistical models for engineers to solve problems in time series analysis. Readers no doubt will learn state-of-the-art techniques from this book.-Hsun-Hsien Chang, Computing Reviews, March 2012My favorite chapters were on dynamic linear models and vector AR and vector ARMA models.-William Seaver, Technometrics, August 2011… a very modern entry to the field of time-series modelling, with a rich reference list of the current lit
Time Series Analysis Forecasting and Control
Box, George E P; Reinsel, Gregory C
2011-01-01
A modernized new edition of one of the most trusted books on time series analysis. Since publication of the first edition in 1970, Time Series Analysis has served as one of the most influential and prominent works on the subject. This new edition maintains its balanced presentation of the tools for modeling and analyzing time series and also introduces the latest developments that have occurred n the field over the past decade through applications from areas such as business, finance, and engineering. The Fourth Edition provides a clearly written exploration of the key methods for building, cl
Visibility Graph Based Time Series Analysis.
Mutua Stephen
Full Text Available Network based time series analysis has made considerable achievements in the recent years. By mapping mono/multivariate time series into networks, one can investigate both it's microscopic and macroscopic behaviors. However, most proposed approaches lead to the construction of static networks consequently providing limited information on evolutionary behaviors. In the present paper we propose a method called visibility graph based time series analysis, in which series segments are mapped to visibility graphs as being descriptions of the corresponding states and the successively occurring states are linked. This procedure converts a time series to a temporal network and at the same time a network of networks. Findings from empirical records for stock markets in USA (S&P500 and Nasdaq and artificial series generated by means of fractional Gaussian motions show that the method can provide us rich information benefiting short-term and long-term predictions. Theoretically, we propose a method to investigate time series from the viewpoint of network of networks.
Forecasting Daily Time Series using Periodic Unobserved Components Time Series Models
Koopman, Siem Jan; Ooms, Marius
2004-01-01
We explore a periodic analysis in the context of unobserved components time series models that decompose time series into components of interest such as trend and seasonal. Periodic time series models allow dynamic characteristics to depend on the period of the year, month, week or day. In the stand
Measuring nonlinear behavior in time series data
Wai, Phoong Seuk; Ismail, Mohd Tahir
2014-12-01
Stationary Test is an important test in detect the time series behavior since financial and economic data series always have missing data, structural change as well as jumps or breaks in the data set. Moreover, stationary test is able to transform the nonlinear time series variable to become stationary by taking difference-stationary process or trend-stationary process. Two different types of hypothesis testing of stationary tests that are Augmented Dickey-Fuller (ADF) test and Kwiatkowski-Philips-Schmidt-Shin (KPSS) test are examine in this paper to describe the properties of the time series variables in financial model. Besides, Least Square method is used in Augmented Dickey-Fuller test to detect the changes of the series and Lagrange multiplier is used in Kwiatkowski-Philips-Schmidt-Shin test to examine the properties of oil price, gold price and Malaysia stock market. Moreover, Quandt-Andrews, Bai-Perron and Chow tests are also use to detect the existence of break in the data series. The monthly index data are ranging from December 1989 until May 2012. Result is shown that these three series exhibit nonlinear properties but are able to transform to stationary series after taking first difference process.
Complex network approach to fractional time series
In order to extract correlation information inherited in stochastic time series, the visibility graph algorithm has been recently proposed, by which a time series can be mapped onto a complex network. We demonstrate that the visibility algorithm is not an appropriate one to study the correlation aspects of a time series. We then employ the horizontal visibility algorithm, as a much simpler one, to map fractional processes onto complex networks. The degree distributions are shown to have parabolic exponential forms with Hurst dependent fitting parameter. Further, we take into account other topological properties such as maximum eigenvalue of the adjacency matrix and the degree assortativity, and show that such topological quantities can also be used to predict the Hurst exponent, with an exception for anti-persistent fractional Gaussian noises. To solve this problem, we take into account the Spearman correlation coefficient between nodes' degrees and their corresponding data values in the original time series
Advanced spectral methods for climatic time series
Ghil, M.; Allen, M.R.; Dettinger, M.D.; Ide, K.; Kondrashov, D.; Mann, M.E.; Robertson, A.W.; Saunders, A.; Tian, Y.; Varadi, F.; Yiou, P.
2002-01-01
The analysis of univariate or multivariate time series provides crucial information to describe, understand, and predict climatic variability. The discovery and implementation of a number of novel methods for extracting useful information from time series has recently revitalized this classical field of study. Considerable progress has also been made in interpreting the information so obtained in terms of dynamical systems theory. In this review we describe the connections between time series analysis and nonlinear dynamics, discuss signal- to-noise enhancement, and present some of the novel methods for spectral analysis. The various steps, as well as the advantages and disadvantages of these methods, are illustrated by their application to an important climatic time series, the Southern Oscillation Index. This index captures major features of interannual climate variability and is used extensively in its prediction. Regional and global sea surface temperature data sets are used to illustrate multivariate spectral methods. Open questions and further prospects conclude the review.
Complex network approach to fractional time series
Manshour, Pouya
2015-10-01
In order to extract correlation information inherited in stochastic time series, the visibility graph algorithm has been recently proposed, by which a time series can be mapped onto a complex network. We demonstrate that the visibility algorithm is not an appropriate one to study the correlation aspects of a time series. We then employ the horizontal visibility algorithm, as a much simpler one, to map fractional processes onto complex networks. The degree distributions are shown to have parabolic exponential forms with Hurst dependent fitting parameter. Further, we take into account other topological properties such as maximum eigenvalue of the adjacency matrix and the degree assortativity, and show that such topological quantities can also be used to predict the Hurst exponent, with an exception for anti-persistent fractional Gaussian noises. To solve this problem, we take into account the Spearman correlation coefficient between nodes' degrees and their corresponding data values in the original time series.
Applied time series analysis and innovative computing
Ao, Sio-Iong
2010-01-01
This text is a systematic, state-of-the-art introduction to the use of innovative computing paradigms as an investigative tool for applications in time series analysis. It includes frontier case studies based on recent research.
Detecting nonlinear structure in time series
We describe an approach for evaluating the statistical significance of evidence for nonlinearity in a time series. The formal application of our method requires the careful statement of a null hypothesis which characterizes a candidate linear process, the generation of an ensemble of ''surrogate'' data sets which are similar to the original time series but consistent with the null hypothesis, and the computation of a discriminating statistic for the original and for each of the surrogate data sets. The idea is to test the original time series against the null hypothesis by checking whether the discriminating statistic computed for the original time series differs significantly from the statistics computed for each of the surrogate sets. While some data sets very cleanly exhibit low-dimensional chaos, there are many cases where the evidence is sketchy and difficult to evaluate. We hope to provide a framework within which such claims of nonlinearity can be evaluated. 5 refs., 4 figs
Bayes linear variance adjustment for time series
Wilkinson, Darren J
2008-01-01
This paper exhibits quadratic products of linear combinations of observables which identify the covariance structure underlying the univariate locally linear time series dynamic linear model. The first- and second-order moments for the joint distribution over these observables are given, allowing Bayes linear learning for the underlying covariance structure for the time series model. An example is given which illustrates the methodology and highlights the practical implications of the theory.
FATS: Feature Analysis for Time Series
Nun, Isadora; Sim, Brandon; Zhu, Ming; Dave, Rahul; Castro, Nicolas; Pichara, Karim
2015-01-01
In this paper, we present the FATS (Feature Analysis for Time Series) library. FATS is a Python library which facilitates and standardizes feature extraction for time series data. In particular, we focus on one application: feature extraction for astronomical light curve data, although the library is generalizable for other uses. We detail the methods and features implemented for light curve analysis, and present examples for its usage.
Combination prediction method of chaotic time series
ZHAO DongHua; RUAN Jiong; CAI ZhiJie
2007-01-01
In the present paper, we propose an approach of combination prediction of chaotic time series. The method is based on the adding-weight one-rank local-region method of chaotic time series. The method allows us to define an interval containing a future value with a given probability, which is obtained by studying the prediction error distribution. Its effectiveness is shown with data generated by Logistic map.
Nonlinear time series: semiparametric and nonparametric methods
Gao, Jiti
2007-01-01
Useful in the theoretical and empirical analysis of nonlinear time series data, semiparametric methods have received extensive attention in the economics and statistics communities over the past twenty years. Recent studies show that semiparametric methods and models may be applied to solve dimensionality reduction problems arising from using fully nonparametric models and methods. Answering the call for an up-to-date overview of the latest developments in the field, "Nonlinear Time Series: S...
Introduction to time series analysis and forecasting
Montgomery, Douglas C; Kulahci, Murat
2008-01-01
An accessible introduction to the most current thinking in and practicality of forecasting techniques in the context of time-oriented data. Analyzing time-oriented data and forecasting are among the most important problems that analysts face across many fields, ranging from finance and economics to production operations and the natural sciences. As a result, there is a widespread need for large groups of people in a variety of fields to understand the basic concepts of time series analysis and forecasting. Introduction to Time Series Analysis and Forecasting presents the time series analysis branch of applied statistics as the underlying methodology for developing practical forecasts, and it also bridges the gap between theory and practice by equipping readers with the tools needed to analyze time-oriented data and construct useful, short- to medium-term, statistically based forecasts.
Time series irreversibility: a visibility graph approach
Lacasa, Lucas; Roldán, Édgar; Parrondo, Juan M R; Luque, Bartolo
2011-01-01
We propose a method to measure real-valued time series irreversibility which combines two differ- ent tools: the horizontal visibility algorithm and the Kullback-Leibler divergence. This method maps a time series to a directed network according to a geometric criterion. The degree of irreversibility of the series is then estimated by the Kullback-Leibler divergence (i.e. the distinguishability) between the in and out degree distributions of the associated graph. The method is computationally effi- cient, does not require any ad hoc symbolization process, and naturally takes into account multiple scales. We find that the method correctly distinguishes between reversible and irreversible station- ary time series, including analytical and numerical studies of its performance for: (i) reversible stochastic processes (uncorrelated and Gaussian linearly correlated), (ii) irreversible stochastic pro- cesses (a discrete flashing ratchet in an asymmetric potential), (iii) reversible (conservative) and irreversible (di...
Multiscale entropy analysis of electroseismic time series
L. Guzmán-Vargas; Ramírez-Rojas, A.; Angulo-Brown, F.
2008-01-01
In this work we use the multiscale entropy method to analyse the variability of geo-electric time series monitored in two sites located in Mexico. In our analysis we consider a period of time from January 1995 to December 1995. We systematically calculate the sample entropy of electroseismic time series. Important differences in the entropy profile for several time scales are observed in records from the same station. In particular, a complex behaviour is observed in the vicinity of a
Time Series Forecasting with Missing Values
Shin-Fu Wu
2015-11-01
Full Text Available Time series prediction has become more popular in various kinds of applications such as weather prediction, control engineering, financial analysis, industrial monitoring, etc. To deal with real-world problems, we are often faced with missing values in the data due to sensor malfunctions or human errors. Traditionally, the missing values are simply omitted or replaced by means of imputation methods. However, omitting those missing values may cause temporal discontinuity. Imputation methods, on the other hand, may alter the original time series. In this study, we propose a novel forecasting method based on least squares support vector machine (LSSVM. We employ the input patterns with the temporal information which is defined as local time index (LTI. Time series data as well as local time indexes are fed to LSSVM for doing forecasting without imputation. We compare the forecasting performance of our method with other imputation methods. Experimental results show that the proposed method is promising and is worth further investigations.
Testing time symmetry in time series using data compression dictionaries
Kennel, Matthew B.
2004-01-01
Time symmetry, often called statistical time reversibility, in a dynamical process means that any segment of time-series output has the same probability of occurrence in the process as its time reversal. A technique, based on symbolic dynamics, is proposed to distinguish such symmetrical processes from asymmetrical ones, given a time-series observation of the otherwise unknown process. Because linear stochastic Gaussian processes, and static nonlinear transformations of them, are statisticall...
Feature Matching in Time Series Modelling
Xia, Yingcun
2011-01-01
Using a time series model to mimic an observed time series has a long history. However, with regard to this objective, conventional estimation methods for discrete-time dynamical models are frequently found to be wanting. In the absence of a true model, we prefer an alternative approach to conventional model fitting that typically involves one-step-ahead prediction errors. Our primary aim is to match the joint probability distribution of the observable time series, including long-term features of the dynamics that underpin the data, such as cycles, long memory and others, rather than short-term prediction. For want of a better name, we call this specific aim {\\it feature matching}. The challenges of model mis-specification, measurement errors and the scarcity of data are forever present in real time series modelling. In this paper, by synthesizing earlier attempts into an extended-likelihood, we develop a systematic approach to empirical time series analysis to address these challenges and to aim at achieving...
Introduction to time series analysis and forecasting
Montgomery, Douglas C; Kulahci, Murat
2015-01-01
Praise for the First Edition ""…[t]he book is great for readers who need to apply the methods and models presented but have little background in mathematics and statistics."" -MAA Reviews Thoroughly updated throughout, Introduction to Time Series Analysis and Forecasting, Second Edition presents the underlying theories of time series analysis that are needed to analyze time-oriented data and construct real-world short- to medium-term statistical forecasts. Authored by highly-experienced academics and professionals in engineering statistics, the Second Edition features discussions on both
Building Chaotic Model From Incomplete Time Series
Siek, Michael; Solomatine, Dimitri
2010-05-01
This paper presents a number of novel techniques for building a predictive chaotic model from incomplete time series. A predictive chaotic model is built by reconstructing the time-delayed phase space from observed time series and the prediction is made by a global model or adaptive local models based on the dynamical neighbors found in the reconstructed phase space. In general, the building of any data-driven models depends on the completeness and quality of the data itself. However, the completeness of the data availability can not always be guaranteed since the measurement or data transmission is intermittently not working properly due to some reasons. We propose two main solutions dealing with incomplete time series: using imputing and non-imputing methods. For imputing methods, we utilized the interpolation methods (weighted sum of linear interpolations, Bayesian principle component analysis and cubic spline interpolation) and predictive models (neural network, kernel machine, chaotic model) for estimating the missing values. After imputing the missing values, the phase space reconstruction and chaotic model prediction are executed as a standard procedure. For non-imputing methods, we reconstructed the time-delayed phase space from observed time series with missing values. This reconstruction results in non-continuous trajectories. However, the local model prediction can still be made from the other dynamical neighbors reconstructed from non-missing values. We implemented and tested these methods to construct a chaotic model for predicting storm surges at Hoek van Holland as the entrance of Rotterdam Port. The hourly surge time series is available for duration of 1990-1996. For measuring the performance of the proposed methods, a synthetic time series with missing values generated by a particular random variable to the original (complete) time series is utilized. There exist two main performance measures used in this work: (1) error measures between the actual
Fractal and natural time analysis of geoelectrical time series
Ramirez Rojas, A.; Moreno-Torres, L. R.; Cervantes, F.
2013-05-01
In this work we show the analysis of geoelectric time series linked with two earthquakes of M=6.6 and M=7.4. That time series were monitored at the South Pacific Mexican coast, which is the most important active seismic subduction zone in México. The geolectric time series were analyzed by using two complementary methods: a fractal analysis, by means of the detrended fluctuation analysis (DFA) in the conventional time, and the power spectrum defined in natural time domain (NTD). In conventional time we found long-range correlations prior to the EQ-occurrences and simultaneously in NTD, the behavior of the power spectrum suggest the possible existence of seismo electric signals (SES) similar with the previously reported in equivalent time series monitored in Greece prior to earthquakes of relevant magnitude.
Layered Ensemble Architecture for Time Series Forecasting.
Rahman, Md Mustafizur; Islam, Md Monirul; Murase, Kazuyuki; Yao, Xin
2016-01-01
Time series forecasting (TSF) has been widely used in many application areas such as science, engineering, and finance. The phenomena generating time series are usually unknown and information available for forecasting is only limited to the past values of the series. It is, therefore, necessary to use an appropriate number of past values, termed lag, for forecasting. This paper proposes a layered ensemble architecture (LEA) for TSF problems. Our LEA consists of two layers, each of which uses an ensemble of multilayer perceptron (MLP) networks. While the first ensemble layer tries to find an appropriate lag, the second ensemble layer employs the obtained lag for forecasting. Unlike most previous work on TSF, the proposed architecture considers both accuracy and diversity of the individual networks in constructing an ensemble. LEA trains different networks in the ensemble by using different training sets with an aim of maintaining diversity among the networks. However, it uses the appropriate lag and combines the best trained networks to construct the ensemble. This indicates LEAs emphasis on accuracy of the networks. The proposed architecture has been tested extensively on time series data of neural network (NN)3 and NN5 competitions. It has also been tested on several standard benchmark time series data. In terms of forecasting accuracy, our experimental results have revealed clearly that LEA is better than other ensemble and nonensemble methods. PMID:25751882
Climate Time Series Analysis and Forecasting
Young, P. C.; Fildes, R.
2009-04-01
This paper will discuss various aspects of climate time series data analysis, modelling and forecasting being carried out at Lancaster. This will include state-dependent parameter, nonlinear, stochastic modelling of globally averaged atmospheric carbon dioxide; the computation of emission strategies based on modern control theory; and extrapolative time series benchmark forecasts of annual average temperature, both global and local. The key to the forecasting evaluation will be the iterative estimation of forecast error based on rolling origin comparisons, as recommended in the forecasting research literature. The presentation will conclude with with a comparison of the time series forecasts with forecasts produced from global circulation models and a discussion of the implications for climate modelling research.
CALENDAR EFFECTS IN MONTHLY TIME SERIES MODELS
Gerhard THURY; Mi ZHOU
2005-01-01
It is not unusual for the level of a monthly economic time series, such as industrial production,retail and wholesale sales, monetary aggregates, telephone calls or road accidents, to be influenced by calendar effects. Such effects arise when changes occur in the level of activity resulting from differences in the composition of calendar between years. The two main sources of calendar effects are trading day variations and moving festivals. Ignoring such calendar effects will lead to substantial distortions in the identification stage of time series modeling. Therefore, it is mandatory to introduce calendar effects, when they are present in a time series, as the component of the model which one wants to estimate.
Fuzzy Information Granules in Time Series Data
HEIKO HOFER; ORTOLANI M; DAVID PATTERSON; FRANK HOEPPNER; ONDINE CALLAN; Berthold, Michael R
2004-01-01
Often, it is desirable to represent a set of time series through typical shapes in order to detect common patterns. The algorithm presented here compares pieces of a different time series in order to find such similar shapes. The use of a fuzzy clustering technique based on fuzzy c-means allows us to detect shapes that belong to a certain group of typical shapes with a degree of membership. Modifications to the original algorithm also allow this matching to be invariant with respect to a scal...
Case study in time series analysis
Zhongjie, Xie
1993-01-01
This book is a monograph on case studies using time series analysis, which includes the main research works applied to practical projects by the author in the past 15 years. The works cover different problems in broad fields, such as: engineering, labour protection, astronomy, physiology, endocrinology, oil development, etc. The first part of this book introduces some basic knowledge of time series analysis which is necessary for the reader to understand the methods and the theory used in the procedure for solving problems. The second part is the main part of this book - case studies in differ
Improving the prediction of chaotic time series
李克平; 高自友; 陈天仑
2003-01-01
One of the features of deterministic chaos is sensitive to initial conditions. This feature limits the prediction horizons of many chaotic systems. In this paper, we propose a new prediction technique for chaotic time series. In our method, some neighbouring points of the predicted point, for which the corresponding local Lyapunov exponent is particularly large, would be discarded during estimating the local dynamics, and thus the error accumulated by the prediction algorithm is reduced. The model is tested for the convection amplitude of Lorenz systems. The simulation results indicate that the prediction technique can improve the prediction of chaotic time series.
Lecture notes for Advanced Time Series Analysis
Madsen, Henrik; Holst, Jan
1997-01-01
A first version of this notes was used at the lectures in Grenoble, and they are now extended and improved (together with Jan Holst), and used in Ph.D. courses on Advanced Time Series Analysis at IMM and at the Department of Mathematical Statistics, University of Lund, 1994, 1997, ......A first version of this notes was used at the lectures in Grenoble, and they are now extended and improved (together with Jan Holst), and used in Ph.D. courses on Advanced Time Series Analysis at IMM and at the Department of Mathematical Statistics, University of Lund, 1994, 1997, ...
Dynamical networks reconstructed from time series
Levnajić, Zoran
2012-01-01
Novel method of reconstructing dynamical networks from empirically measured time series is proposed. By statistically examining the correlations between motions displayed by network nodes, we derive a simple equation that directly yields the adjacency matrix, assuming the intra-network interaction functions to be known. We illustrate the method's implementation on a simple example and discuss the dependence of the reconstruction precision on the properties of time series. Our method is applicable to any network, allowing for reconstruction precision to be maximized, and errors to be estimated.
Introduction to time series and forecasting
Brockwell, Peter J
2016-01-01
This book is aimed at the reader who wishes to gain a working knowledge of time series and forecasting methods as applied to economics, engineering and the natural and social sciences. It assumes knowledge only of basic calculus, matrix algebra and elementary statistics. This third edition contains detailed instructions for the use of the professional version of the Windows-based computer package ITSM2000, now available as a free download from the Springer Extras website. The logic and tools of time series model-building are developed in detail. Numerous exercises are included and the software can be used to analyze and forecast data sets of the user's own choosing. The book can also be used in conjunction with other time series packages such as those included in R. The programs in ITSM2000 however are menu-driven and can be used with minimal investment of time in the computational details. The core of the book covers stationary processes, ARMA and ARIMA processes, multivariate time series and state-space mod...
Time series tapering for short data samples
Kaimal, J.C.; Kristensen, L.
We explore the effect of applying tapered windows on atmospheric data to eliminate overestimation inherent in spectra computed from short time series. Some windows are more effective than others in correcting this distortion. The Hamming window gave the best results with experimental data. The Ha...
Asymptotic spectral theory for nonlinear time series
Shao, Xiaofeng; Wu, Wei Biao
2007-01-01
We consider asymptotic problems in spectral analysis of stationary causal processes. Limiting distributions of periodograms and smoothed periodogram spectral density estimates are obtained and applications to the spectral domain bootstrap are given. Instead of the commonly used strong mixing conditions, in our asymptotic spectral theory we impose conditions only involving (conditional) moments, which are easily verifiable for a variety of nonlinear time series.
On Bayesian Nonparametric Continuous Time Series Models
Karabatsos, George; Walker, Stephen G.
2013-01-01
This paper is a note on the use of Bayesian nonparametric mixture models for continuous time series. We identify a key requirement for such models, and then establish that there is a single type of model which meets this requirement. As it turns out, the model is well known in multiple change-point problems.
Nonlinear Time Series Analysis via Neural Networks
Volná, Eva; Janošek, Michal; Kocian, Václav; Kotyrba, Martin
This article deals with a time series analysis based on neural networks in order to make an effective forex market [Moore and Roche, J. Int. Econ. 58, 387-411 (2002)] pattern recognition. Our goal is to find and recognize important patterns which repeatedly appear in the market history to adapt our trading system behaviour based on them.
Nonlinear time-series analysis revisited.
Bradley, Elizabeth; Kantz, Holger
2015-09-01
In 1980 and 1981, two pioneering papers laid the foundation for what became known as nonlinear time-series analysis: the analysis of observed data-typically univariate-via dynamical systems theory. Based on the concept of state-space reconstruction, this set of methods allows us to compute characteristic quantities such as Lyapunov exponents and fractal dimensions, to predict the future course of the time series, and even to reconstruct the equations of motion in some cases. In practice, however, there are a number of issues that restrict the power of this approach: whether the signal accurately and thoroughly samples the dynamics, for instance, and whether it contains noise. Moreover, the numerical algorithms that we use to instantiate these ideas are not perfect; they involve approximations, scale parameters, and finite-precision arithmetic, among other things. Even so, nonlinear time-series analysis has been used to great advantage on thousands of real and synthetic data sets from a wide variety of systems ranging from roulette wheels to lasers to the human heart. Even in cases where the data do not meet the mathematical or algorithmic requirements to assure full topological conjugacy, the results of nonlinear time-series analysis can be helpful in understanding, characterizing, and predicting dynamical systems. PMID:26428563
Inferring causality from noisy time series data
Mønster, Dan; Fusaroli, Riccardo; Tylén, Kristian;
2016-01-01
even causality direction in synchronized time-series and in the presence of intermediate coupling. We find that the presence of noise deterministically reduces the level of cross-mapping fidelity, while the convergence rate exhibits higher levels of robustness. Finally, we propose that controlled noise...
Nonlinear time-series analysis revisited
Bradley, Elizabeth; Kantz, Holger
2015-09-01
In 1980 and 1981, two pioneering papers laid the foundation for what became known as nonlinear time-series analysis: the analysis of observed data—typically univariate—via dynamical systems theory. Based on the concept of state-space reconstruction, this set of methods allows us to compute characteristic quantities such as Lyapunov exponents and fractal dimensions, to predict the future course of the time series, and even to reconstruct the equations of motion in some cases. In practice, however, there are a number of issues that restrict the power of this approach: whether the signal accurately and thoroughly samples the dynamics, for instance, and whether it contains noise. Moreover, the numerical algorithms that we use to instantiate these ideas are not perfect; they involve approximations, scale parameters, and finite-precision arithmetic, among other things. Even so, nonlinear time-series analysis has been used to great advantage on thousands of real and synthetic data sets from a wide variety of systems ranging from roulette wheels to lasers to the human heart. Even in cases where the data do not meet the mathematical or algorithmic requirements to assure full topological conjugacy, the results of nonlinear time-series analysis can be helpful in understanding, characterizing, and predicting dynamical systems.
Time Series Prediction Based on Chaotic Attractor
LIKe-Ping; CHENTian-Lun; GAOZi-You
2003-01-01
A new prediction technique is proposed for chaotic time series. The usefulness of the technique is that it can kick off some false neighbor points which are not suitable for the local estimation of the dynamics systems. A time-delayed embedding is used to reconstruct the underlying attractor, and the prediction model is based on the time evolution of the topological neighboring in the phase space. We use a feedforward neural network to approximate the local dominant Lyapunov exponent, and choose the spatial neighbors by the Lyapunov exponent. The model is tested for the Mackey-Glass equation and the convection amplitude of lorenz systems. The results indicate that this prediction technique can improve the prediction of chaotic time series.
Multiscale entropy analysis of electroseismic time series
L. Guzmán-Vargas
2008-08-01
Full Text Available In this work we use the multiscale entropy method to analyse the variability of geo-electric time series monitored in two sites located in Mexico. In our analysis we consider a period of time from January 1995 to December 1995. We systematically calculate the sample entropy of electroseismic time series. Important differences in the entropy profile for several time scales are observed in records from the same station. In particular, a complex behaviour is observed in the vicinity of a M=7.4 EQ occurred on 14 September 1995. Besides, we also compare the changes in the entropy of the original data with their corresponding shuffled version.
Time Series Analysis Using Composite Multiscale Entropy
Kung-Yen Lee
2013-03-01
Full Text Available Multiscale entropy (MSE was recently developed to evaluate the complexity of time series over different time scales. Although the MSE algorithm has been successfully applied in a number of different fields, it encounters a problem in that the statistical reliability of the sample entropy (SampEn of a coarse-grained series is reduced as a time scale factor is increased. Therefore, in this paper, the concept of a composite multiscale entropy (CMSE is introduced to overcome this difficulty. Simulation results on both white noise and 1/f noise show that the CMSE provides higher entropy reliablity than the MSE approach for large time scale factors. On real data analysis, both the MSE and CMSE are applied to extract features from fault bearing vibration signals. Experimental results demonstrate that the proposed CMSE-based feature extractor provides higher separability than the MSE-based feature extractor.
Fractal Analysis On Internet Traffic Time Series
Chong, K B
2002-01-01
Fractal behavior and long-range dependence have been observed in tele-traffic measurement and characterization. In this paper we show results of application of the fractal analysis to internet traffic via various methods. Our result demonstrate that the internet traffic exhibits self-similarity. Time-scale analysis show to be an effective way to characterize the local irregularity. Based on the result of this study, these two Internet time series exhibit fractal characteristic with long-range dependence.
Time series regression studies in environmental epidemiology
Bhaskaran, Krishnan; Gasparrini, Antonio; Hajat, Shakoor; Smeeth, Liam; Armstrong, Ben
2013-01-01
Time series regression studies have been widely used in environmental epidemiology, notably in investigating the short-term associations between exposures such as air pollution, weather variables or pollen, and health outcomes such as mortality, myocardial infarction or disease-specific hospital admissions. Typically, for both exposure and outcome, data are available at regular time intervals (e.g. daily pollution levels and daily mortality counts) and the aim is to explore short-term associa...
TIME SERIES FORECASTING USING NEURAL NETWORKS
BOGDAN OANCEA
2013-05-01
Full Text Available Recent studies have shown the classification and prediction power of the Neural Networks. It has been demonstrated that a NN can approximate any continuous function. Neural networks have been successfully used for forecasting of financial data series. The classical methods used for time series prediction like Box-Jenkins or ARIMA assumes that there is a linear relationship between inputs and outputs. Neural Networks have the advantage that can approximate nonlinear functions. In this paper we compared the performances of different feed forward and recurrent neural networks and training algorithms for predicting the exchange rate EUR/RON and USD/RON. We used data series with daily exchange rates starting from 2005 until 2013.
Time series regression studies in environmental epidemiology.
Bhaskaran, Krishnan; Gasparrini, Antonio; Hajat, Shakoor; Smeeth, Liam; Armstrong, Ben
2013-08-01
Time series regression studies have been widely used in environmental epidemiology, notably in investigating the short-term associations between exposures such as air pollution, weather variables or pollen, and health outcomes such as mortality, myocardial infarction or disease-specific hospital admissions. Typically, for both exposure and outcome, data are available at regular time intervals (e.g. daily pollution levels and daily mortality counts) and the aim is to explore short-term associations between them. In this article, we describe the general features of time series data, and we outline the analysis process, beginning with descriptive analysis, then focusing on issues in time series regression that differ from other regression methods: modelling short-term fluctuations in the presence of seasonal and long-term patterns, dealing with time varying confounding factors and modelling delayed ('lagged') associations between exposure and outcome. We finish with advice on model checking and sensitivity analysis, and some common extensions to the basic model. PMID:23760528
Sliced Inverse Regression for Time Series Analysis
Chen, Li-Sue
1995-11-01
In this thesis, general nonlinear models for time series data are considered. A basic form is x _{t} = f(beta_sp{1} {T}X_{t-1},beta_sp {2}{T}X_{t-1},... , beta_sp{k}{T}X_ {t-1},varepsilon_{t}), where x_{t} is an observed time series data, X_{t } is the first d time lag vector, (x _{t},x_{t-1},... ,x _{t-d-1}), f is an unknown function, beta_{i}'s are unknown vectors, varepsilon_{t }'s are independent distributed. Special cases include AR and TAR models. We investigate the feasibility applying SIR/PHD (Li 1990, 1991) (the sliced inverse regression and principal Hessian methods) in estimating beta _{i}'s. PCA (Principal component analysis) is brought in to check one critical condition for SIR/PHD. Through simulation and a study on 3 well -known data sets of Canadian lynx, U.S. unemployment rate and sunspot numbers, we demonstrate how SIR/PHD can effectively retrieve the interesting low-dimension structures for time series data.
Time Series Analysis Using Geometric Template Matching.
Frank, Jordan; Mannor, Shie; Pineau, Joelle; Precup, Doina
2013-03-01
We present a novel framework for analyzing univariate time series data. At the heart of the approach is a versatile algorithm for measuring the similarity of two segments of time series called geometric template matching (GeTeM). First, we use GeTeM to compute a similarity measure for clustering and nearest-neighbor classification. Next, we present a semi-supervised learning algorithm that uses the similarity measure with hierarchical clustering in order to improve classification performance when unlabeled training data are available. Finally, we present a boosting framework called TDEBOOST, which uses an ensemble of GeTeM classifiers. TDEBOOST augments the traditional boosting approach with an additional step in which the features used as inputs to the classifier are adapted at each step to improve the training error. We empirically evaluate the proposed approaches on several datasets, such as accelerometer data collected from wearable sensors and ECG data. PMID:22641699
Univariate time series forecasting algorithm validation
Ismail, Suzilah; Zakaria, Rohaiza; Muda, Tuan Zalizam Tuan
2014-12-01
Forecasting is a complex process which requires expert tacit knowledge in producing accurate forecast values. This complexity contributes to the gaps between end users and expert. Automating this process by using algorithm can act as a bridge between them. Algorithm is a well-defined rule for solving a problem. In this study a univariate time series forecasting algorithm was developed in JAVA and validated using SPSS and Excel. Two set of simulated data (yearly and non-yearly); several univariate forecasting techniques (i.e. Moving Average, Decomposition, Exponential Smoothing, Time Series Regressions and ARIMA) and recent forecasting process (such as data partition, several error measures, recursive evaluation and etc.) were employed. Successfully, the results of the algorithm tally with the results of SPSS and Excel. This algorithm will not just benefit forecaster but also end users that lacking in depth knowledge of forecasting process.
Multivariate Voronoi Outlier Detection for Time Series
Zwilling, Chris E.; Wang, Michelle Yongmei
2014-01-01
Outlier detection is a primary step in many data mining and analysis applications, including healthcare and medical research. This paper presents a general method to identify outliers in multivariate time series based on a Voronoi diagram, which we call Multivariate Voronoi Outlier Detection (MVOD). The approach copes with outliers in a multivariate framework, via designing and extracting effective attributes or features from the data that can take parametric or nonparametric forms. Voronoi d...
Estimation and Forecasting in Time Series Models
Zhang, Ru
2013-01-01
This dissertation covers several topics in estimation and forecasting in time series models. Chapter one is about estimation and feasible conditional forecasts properties from the predictive regressions, which extends previous results of OLS estimation bias in the predictive regression model by considering predictive regressions with possible zero intercepts, and also allowing the regressor to follow either a stationary AR(1) process or unit root process. The main thrust of this chapter is t...
Bayes analysis of time series with covariates
Volf, Petr
Hradec Králové : Gaudeamus, 2005 - (Skalská, H.), s. 421-426 ISBN 978-80-7041-535-1. [Mathematical Methods in Economics 2005 /23./. Hradec Králové (CZ), 14.09.2005-16.09.2005] R&D Projects: GA ČR GA402/04/1294 Institutional research plan: CEZ:AV0Z10750506 Keywords : Bayes analysis * time series * unemployment data Subject RIV: BB - Applied Statistics, Operational Research
Revisiting algorithms for generating surrogate time series
Raeth, C; Papadakis, I E; Brinkmann, W
2011-01-01
The method of surrogates is one of the key concepts of nonlinear data analysis. Here, we demonstrate that commonly used algorithms for generating surrogates often fail to generate truly linear time series. Rather, they create surrogate realizations with Fourier phase correlations leading to non-detections of nonlinearities. We argue that reliable surrogates can only be generated, if one tests separately for static and dynamic nonlinearities.
Applying time series analysis to performance logs
Kubacki, Marcin; Sosnowski, Janusz
2015-09-01
Contemporary computer systems provide mechanisms for monitoring various performance parameters (e.g. processor or memory usage, disc or network transfers), which are collected and stored in performance logs. An important issue is to derive characteristic features describing normal and abnormal behavior of the systems. For this purpose we use various schemes of analyzing time series. They have been adapted to the specificity of performance logs and verified using data collected from real systems. The presented approach is useful in evaluating system dependability.
On clustering fMRI time series
Goutte, Cyril; Toft, Peter Aundal; Rostrup, E.;
1999-01-01
Analysis of fMRI time series is often performed by extracting one or more parameters for the individual voxels. Methods based, e.g., on various statistical tests are then used to yield parameters corresponding to probability of activation or activation strength. However, these methods do not...... between the activation stimulus and the fMRI signal. We present two different clustering algorithms and use them to identify regions of similar activations in an fMRI experiment involving a visual stimulus....
Time-series models in marketing.
Dekimpe, Marnik; Hanssens, DM
2000-01-01
Leeflang and Wittink (2000) identify three past stages in marketing model building and implementation, review the current status, and provide some intriguing thoughts on how the model-building process may evolve in response to ongoing and anticipated developments in the marketing environment. It is interesting to note that time-series techniques are not mentioned in their review of the past, receive considerable attention in their assessment of the current situation (mainly in the context of ...
Nonparametric inference for unbalance time series data
Oliver Linton
2004-01-01
Estimation of heteroskedasticity and autocorrelation consistent covariance matrices (HACs) is a well established problem in time series. Results have been established under a variety of weak conditions on temporal dependence and heterogeneity that allow one to conduct inference on a variety of statistics, see Newey and West (1987), Hansen (1992), de Jong and Davidson (2000), and Robinson (2004). Indeed there is an extensive literature on automating these procedures starting with Andrews (1991...
Nonparametric inference for unbalanced time series data
Linton, Oliver Bruce
2004-01-01
Estimation of heteroskedasticity and autocorrelation consistent covariance matrices (HACs) is a well established problem in time series. Results have been established under a variety of weak conditions on temporal dependence and heterogeneity that allow one to conduct inference on a variety of statistics, see Newey and West (1987), Hansen (1992), de Jong and Davidson (2000), and Robinson (2004). Indeed there is an extensive literature on automating these procedures starting with Andrews (1991...
Evolving time series forecasting ARMA models
Cortez, Paulo; Rocha, Miguel
2004-01-01
Nowadays, the ability to forecast the future, based only on past data, leads to strategic advantages, which may be the key to success in organizations. Time Series Forecasting (TSF) allows the modeling of complex systems as ``black-boxes'', being a focus of attention in several research arenas such as Operational Research, Statistics or Computer Science. Alternative TSF approaches emerged from the Artificial Intelligence arena, where optimization algorithms inspired on natural selection pr...
Time Series Forecasting with Missing Values
Shin-Fu Wu; Chia-Yung Chang; Shie-Jue Lee
2015-01-01
Time series prediction has become more popular in various kinds of applications such as weather prediction, control engineering, financial analysis, industrial monitoring, etc. To deal with real-world problems, we are often faced with missing values in the data due to sensor malfunctions or human errors. Traditionally, the missing values are simply omitted or replaced by means of imputation methods. However, omitting those missing values may cause temporal discontinuity. Imputation methods, o...
Analysis of Polyphonic Musical Time Series
Sommer, Katrin; Weihs, Claus
A general model for pitch tracking of polyphonic musical time series will be introduced. Based on a model of Davy and Godsill (Bayesian harmonic models for musical pitch estimation and analysis, Technical Report 431, Cambridge University Engineering Department, 2002) Davy and Godsill (2002) the different pitches of the musical sound are estimated with MCMC methods simultaneously. Additionally a preprocessing step is designed to improve the estimation of the fundamental frequencies (A comparative study on polyphonic musical time series using MCMC methods. In C. Preisach et al., editors, Data Analysis, Machine Learning, and Applications, Springer, Berlin, 2008). The preprocessing step compares real audio data with an alphabet constructed from the McGill Master Samples (Opolko and Wapnick, McGill University Master Samples [Compact disc], McGill University, Montreal, 1987) and consists of tones of different instruments. The tones with minimal Itakura-Saito distortion (Gray et al., Transactions on Acoustics, Speech, and Signal Processing ASSP-28(4):367-376, 1980) are chosen as first estimates and as starting points for the MCMC algorithms. Furthermore the implementation of the alphabet is an approach for the recognition of the instruments generating the musical time series. Results are presented for mixed monophonic data from McGill and for self recorded polyphonic audio data.
Time Series Forecasting: A Multivariate Stochastic Approach
Sello, Stefano
1999-01-01
This note deals with a multivariate stochastic approach to forecast the behaviour of a cyclic time series. Particular attention is devoted to the problem of the prediction of time behaviour of sunspot numbers for the current 23th cycle. The idea is to consider the previous known n cycles as n particular realizations of a given stochastic process. The aim is to predict the future behaviour of the current n+1th realization given a portion of the curve and the structure of the previous n realiza...
Normalizing the causality between time series
Liang, X San
2015-01-01
Recently, a rigorous yet concise formula has been derived to evaluate the information flow, and hence the causality in a quantitative sense, between time series. To assess the importance of a resulting causality, it needs to be normalized. The normalization is achieved through distinguishing three types of fundamental mechanisms that govern the marginal entropy change of the flow recipient. A normalized or relative flow measures its importance relative to other mechanisms. In analyzing realistic series, both absolute and relative information flows need to be taken into account, since the normalizers for a pair of reverse flows belong to two different entropy balances; it is quite normal that two identical flows may differ a lot in relative importance in their respective balances. We have reproduced these results with several autoregressive models. We have also shown applications to a climate change problem and a financial analysis problem. For the former, reconfirmed is the role of the Indian Ocean Dipole as ...
Argos: An Optimized Time-Series Photometer
Anjum S. Mukadam; R. E. Nather
2005-06-01
We designed a prime focus CCD photometer, Argos, optimized for high speed time-series measurements of blue variables (Nather & Mukadam 2004) for the 2.1 m telescope at McDonald Observatory. Lack of any intervening optics between the primary mirror and the CCD makes the instrument highly efficient.We measure an improvement in sensitivity by a factor of nine over the 3-channel PMT photometers used on the same telescope and for the same exposure time. The CCD frame transfer operation triggered by GPS synchronized pulses serves as an electronic shutter for the photometer. This minimizes the dead time between exposures, but more importantly, allows a precise control of the start and duration of the exposure. We expect the uncertainty in our timing to be less than 100 s.
Directed networks with underlying time structures from multivariate time series
Tanizawa, Toshihiro; Taya, Fumihiko
2014-01-01
In this paper we propose a method of constructing directed networks of time-dependent phenomena from multivariate time series. As the construction method is based on the linear model, the network fully reflects dynamical features of the system such as time structures of periodicities. Furthermore, this method can construct networks even if these time series show no similarity: situations in which common methods fail. We explicitly introduce a case where common methods do not work. This fact indicates the importance of constructing networks based on dynamical perspective, when we consider time-dependent phenomena. We apply the method to multichannel electroencephalography~(EEG) data and the result reveals underlying interdependency among the components in the brain system.
Fractal fluctuations in cardiac time series
West, B. J.; Zhang, R.; Sanders, A. W.; Miniyar, S.; Zuckerman, J. H.; Levine, B. D.; Blomqvist, C. G. (Principal Investigator)
1999-01-01
Human heart rate, controlled by complex feedback mechanisms, is a vital index of systematic circulation. However, it has been shown that beat-to-beat values of heart rate fluctuate continually over a wide range of time scales. Herein we use the relative dispersion, the ratio of the standard deviation to the mean, to show, by systematically aggregating the data, that the correlation in the beat-to-beat cardiac time series is a modulated inverse power law. This scaling property indicates the existence of long-time memory in the underlying cardiac control process and supports the conclusion that heart rate variability is a temporal fractal. We argue that the cardiac control system has allometric properties that enable it to respond to a dynamical environment through scaling.
Time Series Photometry of KZ Lacertae
Joner, Michael D.
2016-01-01
We present BVRI time series photometry of the high amplitude delta Scuti star KZ Lacertae secured using the 0.9-meter telescope located at the Brigham Young University West Mountain Observatory. In addition to the multicolor light curves that are presented, the V data from the last six years of observations are used to plot an O-C diagram in order to determine the ephemeris and evaluate evidence for period change. We wish to thank the Brigham Young University College of Physical and Mathematical Sciences as well as the Department of Physics and Astronomy for their continued support of the research activities at the West Mountain Observatory.
Fourier analysis of time series an introduction
Bloomfield, Peter
2000-01-01
A new, revised edition of a yet unrivaled work on frequency domain analysis Long recognized for his unique focus on frequency domain methods for the analysis of time series data as well as for his applied, easy-to-understand approach, Peter Bloomfield brings his well-known 1976 work thoroughly up to date. With a minimum of mathematics and an engaging, highly rewarding style, Bloomfield provides in-depth discussions of harmonic regression, harmonic analysis, complex demodulation, and spectrum analysis. All methods are clearly illustrated using examples of specific data sets, while ample
Forecasting with nonlinear time series models
Kock, Anders Bredahl; Teräsvirta, Timo
In this paper, nonlinear models are restricted to mean nonlinear parametric models. Several such models popular in time series econo- metrics are presented and some of their properties discussed. This in- cludes two models based on universal approximators: the Kolmogorov- Gabor polynomial model a......- ular case where the data-generating process is a simple artificial neural network model. Suggestions for further reading conclude the paper....... and two versions of a simple artificial neural network model. Techniques for generating multi-period forecasts from nonlinear models recursively are considered, and the direct (non-recursive) method for this purpose is mentioned as well. Forecasting with com- plex dynamic systems, albeit less frequently...
Time series analysis of temporal networks
Sikdar, Sandipan; Ganguly, Niloy; Mukherjee, Animesh
2016-01-01
A common but an important feature of all real-world networks is that they are temporal in nature, i.e., the network structure changes over time. Due to this dynamic nature, it becomes difficult to propose suitable growth models that can explain the various important characteristic properties of these networks. In fact, in many application oriented studies only knowing these properties is sufficient. For instance, if one wishes to launch a targeted attack on a network, this can be done even without the knowledge of the full network structure; rather an estimate of some of the properties is sufficient enough to launch the attack. We, in this paper show that even if the network structure at a future time point is not available one can still manage to estimate its properties. We propose a novel method to map a temporal network to a set of time series instances, analyze them and using a standard forecast model of time series, try to predict the properties of a temporal network at a later time instance. To our aim, we consider eight properties such as number of active nodes, average degree, clustering coefficient etc. and apply our prediction framework on them. We mainly focus on the temporal network of human face-to-face contacts and observe that it represents a stochastic process with memory that can be modeled as Auto-Regressive-Integrated-Moving-Average (ARIMA). We use cross validation techniques to find the percentage accuracy of our predictions. An important observation is that the frequency domain properties of the time series obtained from spectrogram analysis could be used to refine the prediction framework by identifying beforehand the cases where the error in prediction is likely to be high. This leads to an improvement of 7.96% (for error level ≤20%) in prediction accuracy on an average across all datasets. As an application we show how such prediction scheme can be used to launch targeted attacks on temporal networks. Contribution to the Topical Issue
Time series modelling of surface pressure data
Al-Awadhi, Shafeeqah; Jolliffe, Ian
1998-03-01
In this paper we examine time series modelling of surface pressure data, as measured by a barograph, at Herne Bay, England, during the years 1981-1989. Autoregressive moving average (ARMA) models have been popular in many fields over the past 20 years, although applications in climatology have been rather less widespread than in some disciplines. Some recent examples are Milionis and Davies (Int. J. Climatol., 14, 569-579) and Seleshi et al. (Int. J. Climatol., 14, 911-923). We fit standard ARMA models to the pressure data separately for each of six 2-month natural seasons. Differences between the best fitting models for different seasons are discussed. Barograph data are recorded continuously, whereas ARMA models are fitted to discretely recorded data. The effect of different spacings between the fitted data on the models chosen is discussed briefly.Often, ARMA models can give a parsimonious and interpretable representation of a time series, but for many series the assumptions underlying such models are not fully satisfied, and more complex models may be considered. A specific feature of surface pressure data in the UK is that its behaviour is different at high and at low pressures: day-to-day changes are typically larger at low pressure levels than at higher levels. This means that standard assumptions used in fitting ARMA models are not valid, and two ways of overcoming this problem are investigated. Transformation of the data to better satisfy the usual assumptions is considered, as is the use of non-linear, specifically threshold autoregressive (TAR), models.
Ensemble vs. time averages in financial time series analysis
Seemann, Lars; Hua, Jia-Chen; McCauley, Joseph L.; Gunaratne, Gemunu H.
2012-12-01
Empirical analysis of financial time series suggests that the underlying stochastic dynamics are not only non-stationary, but also exhibit non-stationary increments. However, financial time series are commonly analyzed using the sliding interval technique that assumes stationary increments. We propose an alternative approach that is based on an ensemble over trading days. To determine the effects of time averaging techniques on analysis outcomes, we create an intraday activity model that exhibits periodic variable diffusion dynamics and we assess the model data using both ensemble and time averaging techniques. We find that ensemble averaging techniques detect the underlying dynamics correctly, whereas sliding intervals approaches fail. As many traded assets exhibit characteristic intraday volatility patterns, our work implies that ensemble averages approaches will yield new insight into the study of financial markets’ dynamics.
Nonlinear Time Series Analysis Since 1990:Some Personal Reflections
Howel Tong
2002-01-01
I reflect upon the development of nonlinear time series analysis since 1990 by focusing on five major areas of development. These areas include the interface between nonlinear time series analysis and chaos, the nonparametric/semiparametric approach, nonlinear state space modelling, financial time series and nonlinear modelling of panels of time series.
Periodograms for multiband astronomical time series
Ivezic, Z.; VanderPlas, J. T.
2016-05-01
We summarize the multiband periodogram, a general extension of the well-known Lomb-Scargle approach for detecting periodic signals in time- domain data developed by VanderPlas & Ivezic (2015). A Python implementation of this method is available on GitHub. The multiband periodogram significantly improves period finding for randomly sampled multiband light curves (e.g., Pan-STARRS, DES, and LSST), and can treat non-uniform sampling and heteroscedastic errors. The light curves in each band are modeled as arbitrary truncated Fourier series, with the period and phase shared across all bands. The key aspect is the use of Tikhonov regularization which drives most of the variability into the so-called base model common to all bands, while fits for individual bands describe residuals relative to the base model and typically require lower-order Fourier series. We use simulated light curves and randomly subsampled SDSS Stripe 82 data to demonstrate the superiority of this method compared to other methods from the literature, and find that this method will be able to efficiently determine the correct period in the majority of LSST's bright RR Lyrae stars with as little as six months of LSST data.
Periodograms for Multiband Astronomical Time Series
VanderPlas, Jacob T
2015-01-01
This paper introduces the multiband periodogram, a general extension of the well-known Lomb-Scargle approach for detecting periodic signals in time-domain data. In addition to advantages of the Lomb-Scargle method such as treatment of non-uniform sampling and heteroscedastic errors, the multiband periodogram significantly improves period finding for randomly sampled multiband light curves (e.g., Pan-STARRS, DES and LSST). The light curves in each band are modeled as arbitrary truncated Fourier series, with the period and phase shared across all bands. The key aspect is the use of Tikhonov regularization which drives most of the variability into the so-called base model common to all bands, while fits for individual bands describe residuals relative to the base model and typically require lower-order Fourier series. This decrease in the effective model complexity is the main reason for improved performance. We use simulated light curves and randomly subsampled SDSS Stripe 82 data to demonstrate the superiority...
Load Forecasting Using Time Series Models
Mahendran Shitan
2009-09-01
Full Text Available Load forecasting is a process of predicting the future load demands. It is important for power systemplanners and demand controllers in ensuring that there would be enough generation to cope withthe increasing demand. Accurate model for load forecasting can lead to a better budget planning,maintenance scheduling and fuel management. This paper presents an attempt to forecast the maximumdemand of electricity by finding an appropriate time series model. The methods considered in this studyinclude the Naïve method, Exponential smoothing, Seasonal Holt-Winters, ARMA, ARAR algorithm, andRegression with ARMA Errors. The performance of these different methods was evaluated by using theforecasting accuracy criteria namely, the Mean Absolute Error (MAE, Root Mean Square Error (RMSE andMean Absolute Relative Percentage Error (MARPE. Based on these three criteria the pure autoregressivemodel with an order 2, or AR (2 under ARMA family emerged as the best model for forecasting electricitydemand.
Correlation filtering in financial time series
Aste, T; Tumminello, M; Mantegna, R N
2005-01-01
We apply a method to filter relevant information from the correlation coefficient matrix by extracting a network of relevant interactions. This method succeeds to generate networks with the same hierarchical structure of the Minimum Spanning Tree but containing a larger amount of links resulting in a richer network topology allowing loops and cliques. In Tumminello et al. \\cite{TumminielloPNAS05}, we have shown that this method, applied to a financial portfolio of 100 stocks in the USA equity markets, is pretty efficient in filtering relevant information about the clustering of the system and its hierarchical structure both on the whole system and within each cluster. In particular, we have found that triangular loops and 4 element cliques have important and significant relations with the market structure and properties. Here we apply this filtering procedure to the analysis of correlation in two different kind of interest rate time series (16 Eurodollars and 34 US interest rates).
Normalizing the causality between time series
Liang, X. San
2015-08-01
Recently, a rigorous yet concise formula was derived to evaluate information flow, and hence the causality in a quantitative sense, between time series. To assess the importance of a resulting causality, it needs to be normalized. The normalization is achieved through distinguishing a Lyapunov exponent-like, one-dimensional phase-space stretching rate and a noise-to-signal ratio from the rate of information flow in the balance of the marginal entropy evolution of the flow recipient. It is verified with autoregressive models and applied to a real financial analysis problem. An unusually strong one-way causality is identified from IBM (International Business Machines Corporation) to GE (General Electric Company) in their early era, revealing to us an old story, which has almost faded into oblivion, about "Seven Dwarfs" competing with a giant for the mainframe computer market.
Timing calibration and spectral cleaning of LOFAR time series data
Corstanje, A.; Buitink, S.; Enriquez, J. E.; Falcke, H.; Hörandel, J. R.; Krause, M.; Nelles, A.; Rachen, J. P.; Schellart, P.; Scholten, O.; ter Veen, S.; Thoudam, S.; Trinh, T. N. G.
2016-05-01
We describe a method for spectral cleaning and timing calibration of short time series data of the voltage in individual radio interferometer receivers. It makes use of phase differences in fast Fourier transform (FFT) spectra across antenna pairs. For strong, localized terrestrial sources these are stable over time, while being approximately uniform-random for a sum over many sources or for noise. Using only milliseconds-long datasets, the method finds the strongest interfering transmitters, a first-order solution for relative timing calibrations, and faulty data channels. No knowledge of gain response or quiescent noise levels of the receivers is required. With relatively small data volumes, this approach is suitable for use in an online system monitoring setup for interferometric arrays. We have applied the method to our cosmic-ray data collection, a collection of measurements of short pulses from extensive air showers, recorded by the LOFAR radio telescope. Per air shower, we have collected 2 ms of raw time series data for each receiver. The spectral cleaning has a calculated optimal sensitivity corresponding to a power signal-to-noise ratio of 0.08 (or -11 dB) in a spectral window of 25 kHz, for 2 ms of data in 48 antennas. This is well sufficient for our application. Timing calibration across individual antenna pairs has been performed at 0.4 ns precision; for calibration of signal clocks across stations of 48 antennas the precision is 0.1 ns. Monitoring differences in timing calibration per antenna pair over the course of the period 2011 to 2015 shows a precision of 0.08 ns, which is useful for monitoring and correcting drifts in signal path synchronizations. A cross-check method for timing calibration is presented, using a pulse transmitter carried by a drone flying over the array. Timing precision is similar, 0.3 ns, but is limited by transmitter position measurements, while requiring dedicated flights.
Time series models with a common stochastic variance for analysing economic time series
Koopman, S.J.; C.S. Bos
2002-01-01
This discussion paper led to an article in Statistica Neerlandica (2003). Vol. 57, issue 4, pages 439-469. The linear Gaussian state space model for which the common variance istreated as a stochastic time-varying variable is considered for themodelling of economic time series. The focus of this paper is on thesimultaneous estimation of parameters related to the stochasticprocesses of the mean part and the variance part of the model. Theestimation method is based on maximum likelihood and it ...
Timing calibration and spectral cleaning of LOFAR time series data
Corstanje, A; Enriquez, J E; Falcke, H; Hörandel, J R; Krause, M; Nelles, A; Rachen, J P; Schellart, P; Scholten, O; ter Veen, S; Thoudam, S; Trinh, T N G
2016-01-01
We describe a method for spectral cleaning and timing calibration of short voltage time series data from individual radio interferometer receivers. It makes use of the phase differences in Fast Fourier Transform (FFT) spectra across antenna pairs. For strong, localized terrestrial sources these are stable over time, while being approximately uniform-random for a sum over many sources or for noise. Using only milliseconds-long datasets, the method finds the strongest interfering transmitters, a first-order solution for relative timing calibrations, and faulty data channels. No knowledge of gain response or quiescent noise levels of the receivers is required. With relatively small data volumes, this approach is suitable for use in an online system monitoring setup for interferometric arrays. We have applied the method to our cosmic-ray data collection, a collection of measurements of short pulses from extensive air showers, recorded by the LOFAR radio telescope. Per air shower, we have collected 2 ms of raw tim...
Time series modeling for syndromic surveillance
Mandl Kenneth D
2003-01-01
Full Text Available Abstract Background Emergency department (ED based syndromic surveillance systems identify abnormally high visit rates that may be an early signal of a bioterrorist attack. For example, an anthrax outbreak might first be detectable as an unusual increase in the number of patients reporting to the ED with respiratory symptoms. Reliably identifying these abnormal visit patterns requires a good understanding of the normal patterns of healthcare usage. Unfortunately, systematic methods for determining the expected number of (ED visits on a particular day have not yet been well established. We present here a generalized methodology for developing models of expected ED visit rates. Methods Using time-series methods, we developed robust models of ED utilization for the purpose of defining expected visit rates. The models were based on nearly a decade of historical data at a major metropolitan academic, tertiary care pediatric emergency department. The historical data were fit using trimmed-mean seasonal models, and additional models were fit with autoregressive integrated moving average (ARIMA residuals to account for recent trends in the data. The detection capabilities of the model were tested with simulated outbreaks. Results Models were built both for overall visits and for respiratory-related visits, classified according to the chief complaint recorded at the beginning of each visit. The mean absolute percentage error of the ARIMA models was 9.37% for overall visits and 27.54% for respiratory visits. A simple detection system based on the ARIMA model of overall visits was able to detect 7-day-long simulated outbreaks of 30 visits per day with 100% sensitivity and 97% specificity. Sensitivity decreased with outbreak size, dropping to 94% for outbreaks of 20 visits per day, and 57% for 10 visits per day, all while maintaining a 97% benchmark specificity. Conclusions Time series methods applied to historical ED utilization data are an important tool
Peat conditions mapping using MODIS time series
Poggio, Laura; Gimona, Alessandro; Bruneau, Patricia; Johnson, Sally; McBride, Andrew; Artz, Rebekka
2016-04-01
Large areas of Scotland are covered in peatlands, providing an important sink of carbon in their near natural state but act as a potential source of gaseous and dissolved carbon emission if not in good conditions. Data on the condition of most peatlands in Scotland are, however, scarce and largely confined to sites under nature protection designations, often biased towards sites in better condition. The best information available at present is derived from labour intensive field-based monitoring of relatively few designated sites (Common Standard Monitoring Dataset). In order to provide a national dataset of peat conditions, the available point information from the CSM data was modelled with morphological features and information derived from MODIS sensor. In particular we used time series of indices describing vegetation greenness (Enhanced Vegetation Index), water availability (Normalised Water Difference index), Land Surface Temperature and vegetation productivity (Gross Primary productivity). A scorpan-kriging approach was used, in particular using Generalised Additive Models for the description of the trend. The model provided the probability of a site to be in favourable conditions and the uncertainty of the predictions was taken into account. The internal validation (leave-one-out) provided a mis-classification error of around 0.25. The derived dataset was then used, among others, in the decision making process for the selection of sites for restoration.
Climate Prediction Center (CPC) Global Temperature Time Series
National Oceanic and Atmospheric Administration, Department of Commerce — The global temperature time series provides time series charts using station based observations of daily temperature. These charts provide information about the...
Climate Prediction Center (CPC) Global Precipitation Time Series
National Oceanic and Atmospheric Administration, Department of Commerce — The global precipitation time series provides time series charts showing observations of daily precipitation as well as accumulated precipitation compared to normal...
Some results of analysis of source position time series
Malkin, Zinovy
2015-01-01
Source position time series produced by International VLBI Service for Geodesy and astrometry (IVS) Analysis Centers were analyzed. These series was computed using different software and analysis strategy. Comparison of this series showed that they have considerably different scatter and systematic behavior. Based on the inspection of all the series, new sources were identified as sources with irregular (non-random) position variations. Two statistics used to estimate the noise level in the time series, namely RMS and ADEV were compared.
Deflation-based separation of uncorrelated stationary time series
Miettinen, Jari; Nordhausen, Klaus; Oja, Hannu; Taskinen, Sara
2014-01-01
In this paper we assume that the observed pp time series are linear combinations of pp latent uncorrelated weakly stationary time series. The problem is then to find an estimate for an unmixing matrix that transforms the observed time series back to uncorrelated time series. The so called SOBI (Second Order Blind Identification) estimate aims at a joint diagonalization of the covariance matrix and several autocovariance matrices with varying lags. In this paper, we propose a novel procedure t...
An introduction to state space time series analysis.
Commandeur, J.J.F. & Koopman, S.J.
2007-01-01
Providing a practical introduction to state space methods as applied to unobserved components time series models, also known as structural time series models, this book introduces time series analysis using state space methodology to readers who are neither familiar with time series analysis, nor with state space methods. The only background required in order to understand the material presented in the book is a basic knowledge of classical linear regression models, of which a brief review is...
Seasonal Time Series Analysis Based on Genetic Algorithm
无
2007-01-01
Pattern discovery from the seasonal time-series is of importance. Traditionally, most of the algorithms of pattern discovery in time series are similar. A novel mode of time series is proposed which integrates the Genetic Algorithm (GA) for the actual problem. The experiments on the electric power yield sequence models show that this algorithm is practicable and effective.
Notes on time serie analysis, ARIMA models and signal extraction
Kaiser, Regina; Maravall, Agustín
2000-01-01
Present practice in applied time series work, mostly at economic policy or data producing agencies, relies heavily on using moving average filters to estimate unobserved components (or signals) in time series, such as the seasonally adjusted series, the trend, or the cycle. The purpose of the present paper is to provide an informal introduction to the time series analysis tools and concepts required by the user or analyst to understand the basic methodology behind the application of filters. ...
Generalized Framework for Similarity Measure of Time Series
Hongsheng Yin
2014-01-01
Full Text Available Currently, there is no definitive and uniform description for the similarity of time series, which results in difficulties for relevant research on this topic. In this paper, we propose a generalized framework to measure the similarity of time series. In this generalized framework, whether the time series is univariable or multivariable, and linear transformed or nonlinear transformed, the similarity of time series is uniformly defined using norms of vectors or matrices. The definitions of the similarity of time series in the original space and the transformed space are proved to be equivalent. Furthermore, we also extend the theory on similarity of univariable time series to multivariable time series. We present some experimental results on published time series datasets tested with the proposed similarity measure function of time series. Through the proofs and experiments, it can be claimed that the similarity measure functions of linear multivariable time series based on the norm distance of covariance matrix and nonlinear multivariable time series based on kernel function are reasonable and practical.
Outlier Detection in Structural Time Series Models
Marczak, Martyna; Proietti, Tommaso
Structural change affects the estimation of economic signals, like the underlying growth rate or the seasonally adjusted series. An important issue, which has attracted a great deal of attention also in the seasonal adjustment literature, is its detection by an expert procedure. The general–to–sp...
Time and ensemble averaging in time series analysis
Latka, Miroslaw; Jernajczyk, Wojciech; West, Bruce J
2010-01-01
In many applications expectation values are calculated by partitioning a single experimental time series into an ensemble of data segments of equal length. Such single trajectory ensemble (STE) is a counterpart to a multiple trajectory ensemble (MTE) used whenever independent measurements or realizations of a stochastic process are available. The equivalence of STE and MTE for stationary systems was postulated by Wang and Uhlenbeck in their classic paper on Brownian motion (Rev. Mod. Phys. 17, 323 (1945)) but surprisingly has not yet been proved. Using the stationary and ergodic paradigm of statistical physics -- the Ornstein-Uhlenbeck (OU) Langevin equation, we revisit Wang and Uhlenbeck's postulate. In particular, we find that the variance of the solution of this equation is different for these two ensembles. While the variance calculated using the MTE quantifies the spreading of independent trajectories originating from the same initial point, the variance for STE measures the spreading of two correlated r...
Hidden Markov Models for Time Series An Introduction Using R
Zucchini, Walter
2009-01-01
Illustrates the flexibility of HMMs as general-purpose models for time series data. This work presents an overview of HMMs for analyzing time series data, from continuous-valued, circular, and multivariate series to binary data, bounded and unbounded counts and categorical observations.
Vyhnalek, Brian; Zurcher, Ulrich; O'Dwyer, Rebecca; Kaufman, Miron
2009-10-01
A wide range of heart rate irregularities have been reported in small studies of patients with temporal lobe epilepsy [TLE]. We hypothesize that patients with TLE display cardiac dysautonomia in either a subclinical or clinical manner. In a small study, we have retrospectively identified (2003-8) two groups of patients from the epilepsy monitoring unit [EMU] at the Cleveland Clinic. No patients were diagnosed with cardiovascular morbidities. The control group consisted of patients with confirmed pseudoseizures and the experimental group had confirmed right temporal lobe epilepsy through a seizure free outcome after temporal lobectomy. We quantified the heart rate variability using the approximate entropy [ApEn]. We found similar values of the ApEn in all three states of consciousness (awake, sleep, and proceeding seizure onset). In the TLE group, there is some evidence for greater variability in the awake than in either the sleep or proceeding seizure onset. Here we present results for mathematically-generated time series: the heart rate fluctuations ξ follow the γ statistics i.e., p(ξ)=γ-1(k) ξ^k exp(-ξ). This probability function has well-known properties and its Shannon entropy can be expressed in terms of the γ-function. The parameter k allows us to generate a family of heart rate time series with different statistics. The ApEn calculated for the generated time series for different values of k mimic the properties found for the TLE and pseudoseizure group. Our results suggest that the ApEn is an effective tool to probe differences in statistics of heart rate fluctuations.
Multiscale entropy to distinguish physiologic and synthetic RR time series.
Costa, M; Goldberger, A L; Peng, C-K
2002-01-01
We address the challenge of distinguishing physiologic interbeat interval time series from those generated by synthetic algorithms via a newly developed multiscale entropy method. Traditional measures of time series complexity only quantify the degree of regularity on a single time scale. However, many physiologic variables, such as heart rate, fluctuate in a very complex manner and present correlations over multiple time scales. We have proposed a new method to calculate multiscale entropy from complex signals. In order to distinguish between physiologic and synthetic time series, we first applied the method to a learning set of RR time series derived from healthy subjects. We empirically established selected criteria characterizing the entropy dependence on scale factor for these datasets. We then applied this algorithm to the CinC 2002 test datasets. Using only the multiscale entropy method, we correctly classified 48 of 50 (96%) time series. In combination with Fourier spectral analysis, we correctly classified all time series. PMID:14686448
Efficient Algorithms for Segmentation of Item-Set Time Series
Chundi, Parvathi; Rosenkrantz, Daniel J.
We propose a special type of time series, which we call an item-set time series, to facilitate the temporal analysis of software version histories, email logs, stock market data, etc. In an item-set time series, each observed data value is a set of discrete items. We formalize the concept of an item-set time series and present efficient algorithms for segmenting a given item-set time series. Segmentation of a time series partitions the time series into a sequence of segments where each segment is constructed by combining consecutive time points of the time series. Each segment is associated with an item set that is computed from the item sets of the time points in that segment, using a function which we call a measure function. We then define a concept called the segment difference, which measures the difference between the item set of a segment and the item sets of the time points in that segment. The segment difference values are required to construct an optimal segmentation of the time series. We describe novel and efficient algorithms to compute segment difference values for each of the measure functions described in the paper. We outline a dynamic programming based scheme to construct an optimal segmentation of the given item-set time series. We use the item-set time series segmentation techniques to analyze the temporal content of three different data sets—Enron email, stock market data, and a synthetic data set. The experimental results show that an optimal segmentation of item-set time series data captures much more temporal content than a segmentation constructed based on the number of time points in each segment, without examining the item set data at the time points, and can be used to analyze different types of temporal data.
TIME SERIES ANALYSIS USING A UNIQUE MODEL OF TRANSFORMATION
Goran Klepac
2007-12-01
Full Text Available REFII1 model is an authorial mathematical model for time series data mining. The main purpose of that model is to automate time series analysis, through a unique transformation model of time series. An advantage of this approach of time series analysis is the linkage of different methods for time series analysis, linking traditional data mining tools in time series, and constructing new algorithms for analyzing time series. It is worth mentioning that REFII model is not a closed system, which means that we have a finite set of methods. At first, this is a model for transformation of values of time series, which prepares data used by different sets of methods based on the same model of transformation in a domain of problem space. REFII model gives a new approach in time series analysis based on a unique model of transformation, which is a base for all kind of time series analysis. The advantage of REFII model is its possible application in many different areas such as finance, medicine, voice recognition, face recognition and text mining.
Multifractal Analysis of Aging and Complexity in Heartbeat Time Series
Muñoz D., Alejandro; Almanza V., Victor H.; del Río C., José L.
2004-09-01
Recently multifractal analysis has been used intensively in the analysis of physiological time series. In this work we apply the multifractal analysis to the study of heartbeat time series from healthy young subjects and other series obtained from old healthy subjects. We show that this multifractal formalism could be a useful tool to discriminate these two kinds of series. We used the algorithm proposed by Chhabra and Jensen that provides a highly accurate, practical and efficient method for the direct computation of the singularity spectrum. Aging causes loss of multifractality in the heartbeat time series, it means that heartbeat time series of elderly persons are less complex than the time series of young persons. This analysis reveals a new level of complexity characterized by the wide range of necessary exponents to characterize the dynamics of young people.
Ruin Probability in Linear Time Series Model
ZHANG Lihong
2005-01-01
This paper analyzes a continuous time risk model with a linear model used to model the claim process. The time is discretized stochastically using the times when claims occur, using Doob's stopping time theorem and martingale inequalities to obtain expressions for the ruin probability as well as both exponential and non-exponential upper bounds for the ruin probability for an infinite time horizon. Numerical results are included to illustrate the accuracy of the non-exponential bound.
Visibility graph network analysis of gold price time series
Long, Yu
2013-08-01
Mapping time series into a visibility graph network, the characteristics of the gold price time series and return temporal series, and the mechanism underlying the gold price fluctuation have been explored from the perspective of complex network theory. The network degree distribution characters, which change from power law to exponent law when the series was shuffled from original sequence, and the average path length characters, which change from L∼lnN into lnL∼lnN as the sequence was shuffled, demonstrate that price series and return series are both long-rang dependent fractal series. The relations of Hurst exponent to the power-law exponent of degree distribution demonstrate that the logarithmic price series is a fractal Brownian series and the logarithmic return series is a fractal Gaussian series. Power-law exponents of degree distribution in a time window changing with window moving demonstrates that a logarithmic gold price series is a multifractal series. The Power-law average clustering coefficient demonstrates that the gold price visibility graph is a hierarchy network. The hierarchy character, in light of the correspondence of graph to price fluctuation, means that gold price fluctuation is a hierarchy structure, which appears to be in agreement with Elliot’s experiential Wave Theory on stock price fluctuation, and the local-rule growth theory of a hierarchy network means that the hierarchy structure of gold price fluctuation originates from persistent, short term factors, such as short term speculation.
On correlations and fractal characteristics of time series
Vitanov, N K; Yankulova, E D; Vitanov, Nikolay K.; Sakai, kenschi; Yankulova, Elka D.
2005-01-01
Correlation analysis is convenient and frequently used tool for investigation of time series from complex systems. Recently new methods such as the multifractal detrended fluctuation analysis (MFDFA) and the wavelet transform modulus maximum method (WTMM) have been developed. By means of these methods (i) we can investigate long-range correlations in time series and (ii) we can calculate fractal spectra of these time series. But opposite to the classical tool for correlation analysis - the autocorrelation function, the newly developed tools are not applicable to all kinds of time series. The unappropriate application of MFDFA or WTMM leads to wrong results and conclusions. In this article we discuss the opportunities and risks connected to the application of the MFDFA method to time series from a random number generator and to experimentally measured time series (i) for accelerations of an agricultural tractor and (ii) for the heartbeat activity of {\\sl Drosophila melanogaster}. Our main goal is to emphasize ...
Non-parametric causal inference for bivariate time series
McCracken, James M
2015-01-01
We introduce new quantities for exploratory causal inference between bivariate time series. The quantities, called penchants and leanings, are computationally straightforward to apply, follow directly from assumptions of probabilistic causality, do not depend on any assumed models for the time series generating process, and do not rely on any embedding procedures; these features may provide a clearer interpretation of the results than those from existing time series causality tools. The penchant and leaning are computed based on a structured method for computing probabilities.
Intrusion Detection Forecasting Using Time Series for Improving Cyber Defence
Abdullah, Azween Bin; Pillai, Thulasyammal Ramiah; Cai, Long Zheng
2015-01-01
The strength of time series modeling is generally not used in almost all current intrusion detection and prevention systems. By having time series models, system administrators will be able to better plan resource allocation and system readiness to defend against malicious activities. In this paper, we address the knowledge gap by investigating the possible inclusion of a statistical based time series modeling that can be seamlessly integrated into existing cyber defense system. Cyber-attack ...
Power-weighted densities for time series data
McCarthy, Daniel M.; Jensen, Shane T.
2016-01-01
While time series prediction is an important, actively studied problem, the predictive accuracy of time series models is complicated by non-stationarity. We develop a fast and effective approach to allow for non-stationarity in the parameters of a chosen time series model. In our power-weighted density (PWD) approach, observations in the distant past are down-weighted in the likelihood function relative to more recent observations, while still giving the practitioner control over the choice o...
Efficient use of correlation entropy for analysing time series data
K P Harikrishnan; R Misra; G Ambika
2009-02-01
The correlation dimension 2 and correlation entropy 2 are both important quantifiers in nonlinear time series analysis. However, use of 2 has been more common compared to 2 as a discriminating measure. One reason for this is that 2 is a static measure and can be easily evaluated from a time series. However, in many cases, especially those involving coloured noise, 2 is regarded as a more useful measure. Here we present an efficient algorithmic scheme to compute 2 directly from a time series data and show that 2 can be used as a more effective measure compared to 2 for analysing practical time series involving coloured noise.
Interpretable Early Classification of Multivariate Time Series
Ghalwash, Mohamed F.
2013-01-01
Recent advances in technology have led to an explosion in data collection over time rather than in a single snapshot. For example, microarray technology allows us to measure gene expression levels in different conditions over time. Such temporal data grants the opportunity for data miners to develop algorithms to address domain-related problems,…
Time Series Properties of Expectation Biases
Kinari, Yusuke
2011-01-01
This study exammes time senes properties of expectation biases usmg a highfrequency survey on stock price forecasts, which required participants to forecast the Nikkei 225 over three forecasting horizons: one day, one week, and one month ahead. Constructing proxies for overconfidence and optimism as the expectation biases, this study shows that overconfidence is likely to remain stable over time while optimism is not. Moreover, a relationship exists between optimism and stock price movement, ...
Studies on time series applications in environmental sciences
Bărbulescu, Alina
2016-01-01
Time series analysis and modelling represent a large study field, implying the approach from the perspective of the time and frequency, with applications in different domains. Modelling hydro-meteorological time series is difficult due to the characteristics of these series, as long range dependence, spatial dependence, the correlation with other series. Continuous spatial data plays an important role in planning, risk assessment and decision making in environmental management. In this context, in this book we present various statistical tests and modelling techniques used for time series analysis, as well as applications to hydro-meteorological series from Dobrogea, a region situated in the south-eastern part of Romania, less studied till now. Part of the results are accompanied by their R code. .
Volatility modeling of rainfall time series
Yusof, Fadhilah; Kane, Ibrahim Lawal
2013-07-01
Networks of rain gauges can provide a better insight into the spatial and temporal variability of rainfall, but they tend to be too widely spaced for accurate estimates. A way to estimate the spatial variability of rainfall between gauge points is to interpolate between them. This paper evaluates the spatial autocorrelation of rainfall data in some locations in Peninsular Malaysia using geostatistical technique. The results give an insight on the spatial variability of rainfall in the area, as such, two rain gauges were selected for an in-depth study of the temporal dependence of the rainfall data-generating process. It could be shown that rainfall data are affected by nonlinear characteristics of the variance often referred to as variance clustering or volatility, where large changes tend to follow large changes and small changes tend to follow small changes. The autocorrelation structure of the residuals and the squared residuals derived from autoregressive integrated moving average (ARIMA) models were inspected, the residuals are uncorrelated but the squared residuals show autocorrelation, and the Ljung-Box test confirmed the results. A test based on the Lagrange multiplier principle was applied to the squared residuals from the ARIMA models. The results of this auxiliary test show a clear evidence to reject the null hypothesis of no autoregressive conditional heteroskedasticity (ARCH) effect. Hence, it indicates that generalized ARCH (GARCH) modeling is necessary. An ARIMA error model is proposed to capture the mean behavior and a GARCH model for modeling heteroskedasticity (variance behavior) of the residuals from the ARIMA model. Therefore, the composite ARIMA-GARCH model captures the dynamics of daily rainfall in the study area. On the other hand, seasonal ARIMA model became a suitable model for the monthly average rainfall series of the same locations treated.
Recovery of the Time-Evolution Equation of Time-Delay Systems from Time Series
Bünner, M J; Kittel, A; Parisi, J; Meyer, Th.
1997-01-01
We present a method for time series analysis of both, scalar and nonscalar time-delay systems. If the dynamics of the system investigated is governed by a time-delay induced instability, the method allows to determine the delay time. In a second step, the time-delay differential equation can be recovered from the time series. The method is a generalization of our recently proposed method suitable for time series analysis of {\\it scalar} time-delay systems. The dynamics is not required to be settled on its attractor, which also makes transient motion accessible to the analysis. If the motion actually takes place on a chaotic attractor, the applicability of the method does not depend on the dimensionality of the chaotic attractor - one main advantage over all time series analysis methods known until now. For demonstration, we analyze time series, which are obtained with the help of the numerical integration of a two-dimensional time-delay differential equation. After having determined the delay time, we recover...
Fitting dynamic fator models to nonstationary time series
Eichler, M.; Motta, Giovanni; Von Sachs, Rainer
2008-01-01
Factor modelling of a large time series panel has widely proven useful to reduce its cross-sectional dimensionality. This is done by explaining common co-movements in the panel through the existence of a small number of common components, up to some idiosyncratic behaviour of each individual series. To capture serial correlation in the common components, a dynamic structure is used as in traditional (uni- or multivariate) time series analysis of second order structure,i.e. allowing f...
Fitting dynamic factor models to non-stationary time series
Eichler Michael; Motta Giovanni; Sachs Rainer von
2009-01-01
Factor modelling of a large time series panel has widely proven useful to reduce its cross-sectional dimensionality. This is done by explaining common co-movements in the panel through the existence of a small number of common components, up to some idiosyncratic behaviour of each individual series. To capture serial correlation in the common components, a dynamic structure is used as in traditional (uni- or multivariate) time series analysis of second order structure, i.e. allowing for infin...
Time series prediction using wavelet process neural network
Ding Gang; Zhong Shi-Sheng; Li Yang
2008-01-01
In the real world, the inputs of many complicated systems are time-varying functions or processes. In order to predict the outputs of these systems with high speed and accuracy, this paper proposes a time series prediction model based on the wavelet process neural network, and develops the corresponding learning algorithm based on the expansion of the orthogonal basis functions. The effectiveness of the proposed time series prediction model and its learning algorithm is proved by the Mackey-Glass time series prediction, and the comparative prediction results indicate that the proposed time series prediction model based on the wavelet process neural network seems to perform well and appears suitable for using as a good tool to predict the highly complex nonlinear time series.
Robust Forecasting of Non-Stationary Time Series
Croux, C.; Fried, R.; Gijbels, I.; Mahieu, K.
2010-01-01
This paper proposes a robust forecasting method for non-stationary time series. The time series is modelled using non-parametric heteroscedastic regression, and fitted by a localized MM-estimator, combining high robustness and large efficiency. The proposed method is shown to produce reliable foreca
A Computer Evolution in Teaching Undergraduate Time Series
Hodgess, Erin M.
2004-01-01
In teaching undergraduate time series courses, we have used a mixture of various statistical packages. We have finally been able to teach all of the applied concepts within one statistical package; R. This article describes the process that we use to conduct a thorough analysis of a time series. An example with a data set is provided. We compare…
Two-fractal overlap time series: Earthquakes and market crashes
Bikas K Chakrabarti; Arnab Chatterjee; Pratip Bhattacharyya
2008-08-01
We find prominent similarities in the features of the time series for the (model earthquakes or) overlap of two Cantor sets when one set moves with uniform relative velocity over the other and time series of stock prices. An anticipation method for some of the crashes have been proposed here, based on these observations.
Stata: The language of choice for time series analysis?
Christopher F. Baum
2004-01-01
This paper discusses the use of Stata for the analysis of time series and panel data. The evolution of time-series capabilities in Stata is reviewed. Facilities for data management, graphics, and econometric analysis from both official Stata and the user community are discussed. A new routine to provide moving-window regression estimates, rollreg, is described, and its use illustrated.
Parameterizing unconditional skewness in models for financial time series
He, Changli; Silvennoinen, Annastiina; Teräsvirta, Timo
In this paper we consider the third-moment structure of a class of time series models. It is often argued that the marginal distribution of financial time series such as returns is skewed. Therefore it is of importance to know what properties a model should possess if it is to accommodate...
Using Time-Series Regression to Predict Academic Library Circulations.
Brooks, Terrence A.
1984-01-01
Four methods were used to forecast monthly circulation totals in 15 midwestern academic libraries: dummy time-series regression, lagged time-series regression, simple average (straight-line forecasting), monthly average (naive forecasting). In tests of forecasting accuracy, dummy regression method and monthly mean method exhibited smallest average…
Two Fractal Overlap Time Series: Earthquakes and Market Crashes
Chakrabarti, Bikas K.; Arnab Chatterjee; Pratip Bhattacharyya
2007-01-01
We find prominent similarities in the features of the time series for the (model earthquakes or) overlap of two Cantor sets when one set moves with uniform relative velocity over the other and time series of stock prices. An anticipation method for some of the crashes have been proposed here, based on these observations.
Reprocessed height time series for GPS stations
S. Rudenko; Schön, N.; Uhlemann, M; G. Gendt
2013-01-01
Precise weekly positions of 403 Global Positioning System (GPS) stations located worldwide are obtained by reprocessing GPS data of these stations for the time span from 4 January 1998 until 29 December 2007. The processing algorithms and models used as well as the solution and results obtained are presented. Vertical velocities of 266 GPS stations having a tracking history longer than 2.5 yr are computed; 107 of them are GPS stations located at tide gauges (TIGA observing stations). The vert...
Using wavelets for time series forecasting: Does it pay off?
Schlüter, Stephan; Deuschle, Carola
2010-01-01
By means of wavelet transform a time series can be decomposed into a time dependent sum of frequency components. As a result we are able to capture seasonalities with time-varying period and intensity, which nourishes the belief that incorporating the wavelet transform in existing forecasting methods can improve their quality. The article aims to verify this by comparing the power of classical and wavelet based techniques on the basis of four time series, each of them having individual charac...
Time-Series Photometric Surveys: Some Musings
Howell, S. B.
We live in the era of large astronomical surveys aimed at collecting high photometric precision, high time resolution, and long term (near) continuous observations. Such surveys discover many variable sources and their study has led to new paradigms in observational astronomy. Periodic variables have a long and venerable history in astronomy being highly useful as distance ladders, to investigate stellar interior physics and to map out Galactic structure. However, typically less than 10% of all variable sources are periodic and a detailed understanding of the majority of variables, the non-periodic sources, is lacking. What can we learn from non-periodic variables? Are there alternative techniques or types of study that may help elucidate their true nature? This talk will attempt to provide a short review of our understanding of variable sources and provide some suggestions for a methodology toward the study of non-variable astronomical sources.
Useful Pattern Mining on Time Series
Goumatianos, Nikitas; Christou, Ioannis T; Lindgren, Peter
2013-01-01
We present the architecture of a “useful pattern” mining system that is capable of detecting thousands of different candlestick sequence patterns at the tick or any higher granularity levels. The system architecture is highly distributed and performs most of its highly compute-intensive aggregation......% or higher increase (or, alternatively, decrease) in a chosen property of the stock (e.g. close-value) within a given time-window (e.g. 5 days). Initial results from a first prototype implementation of the architecture show that after training on a large set of stocks, the system is capable of finding...... a significant number of candlestick sequences whose output signals (measured against an unseen set of stocks) have predictive accuracy which varies between 60% and 95% depended on the type of pattern....
Interactive analysis of gappy bivariate time series using AGSS
Lewis, Peter A.W.; Ray, Bonnie K.
1992-01-01
Bivariate time series which display nonstationary behavior, such as cycles or long-term trends, are common in fields such as oceanography and meteorology. These are usually very large-scale data sets and often may contain long gaps of missing values in one or both series, with the gaps perhaps occurring at different time periods in the two series. We present a simplified but effective method of interactively examining and filling in the missing values in such series using extensions of the me...
Comparison of New and Old Sunspot Number Time Series
Cliver, E. W.
2016-06-01
Four new sunspot number time series have been published in this Topical Issue: a backbone-based group number in Svalgaard and Schatten (Solar Phys., 2016; referred to here as SS, 1610 - present), a group number series in Usoskin et al. (Solar Phys., 2016; UEA, 1749 - present) that employs active day fractions from which it derives an observational threshold in group spot area as a measure of observer merit, a provisional group number series in Cliver and Ling (Solar Phys., 2016; CL, 1841 - 1976) that removed flaws in the Hoyt and Schatten (Solar Phys. 179, 189, 1998a; 181, 491, 1998b) normalization scheme for the original relative group sunspot number ( RG, 1610 - 1995), and a corrected Wolf (international, RI) number in Clette and Lefèvre (Solar Phys., 2016; SN, 1700 - present). Despite quite different construction methods, the four new series agree well after about 1900. Before 1900, however, the UEA time series is lower than SS, CL, and SN, particularly so before about 1885. Overall, the UEA series most closely resembles the original RG series. Comparison of the UEA and SS series with a new solar wind B time series (Owens et al. in J. Geophys. Res., 2016; 1845 - present) indicates that the UEA time series is too low before 1900. We point out incongruities in the Usoskin et al. (Solar Phys., 2016) observer normalization scheme and present evidence that this method under-estimates group counts before 1900. In general, a correction factor time series, obtained by dividing an annual group count series by the corresponding yearly averages of raw group counts for all observers, can be used to assess the reliability of new sunspot number reconstructions.
Application of p-adic analysis to time series
Khrennikov, A. Yu.; Kozyrev, S. V.; Oleschko, K. (collab.); Jaramillo, A. G.; Lopez, M. de Jesus Correa
2013-01-01
Time series defined by a p-adic pseudo-differential equation is investigated using the expansion of the time series over p-adic wavelets. Quadratic correlation function is computed. This correlation function shows a degree--like behavior and is locally constant for some time periods. It is natural to apply this kind of models for the investigation of avalanche processes and punctuated equilibrium as well as fractal-like analysis of time series generated by measurement of pressure in oil wells.
Time series analysis and inverse theory for geophysicists
Junzo Kasahara
2006-01-01
@@ Thanks to the advances in geophysical measurement technologies, most geophysical data are now recorded in digital form. But to extract the ‘Earth's nature’ from observed data, it is necessary to apply the signal-processing method to the time-series data, seismograms and geomagnetic records being the most common. The processing of time-series data is one of the major subjects of this book.By the processing of time series data, numerical values such as travel-times are obtained.The first stage of data analysis is forward modeling, but the more advanced step is the inversion method. This is the second subject of this book.
Performance of multifractal detrended fluctuation analysis on short time series
Lopez, Juan Luis
2013-01-01
The performance of the multifractal detrended analysis on short time series is evaluated for synthetic samples of several mono- and multifractal models. The reconstruction of the generalized Hurst exponents is used to determine the range of applicability of the method and the precision of its results as a function of the decreasing length of the series. As an application the series of the daily exchange rate between the U.S. dollar and the euro is studied.
Database for Hydrological Time Series of Inland Waters (DAHITI)
Schwatke, Christian; Dettmering, Denise
2016-04-01
Satellite altimetry was designed for ocean applications. However, since some years, satellite altimetry is also used over inland water to estimate water level time series of lakes, rivers and wetlands. The resulting water level time series can help to understand the water cycle of system earth and makes altimetry to a very useful instrument for hydrological applications. In this poster, we introduce the "Database for Hydrological Time Series of Inland Waters" (DAHITI). Currently, the database contains about 350 water level time series of lakes, reservoirs, rivers, and wetlands which are freely available after a short registration process via http://dahiti.dgfi.tum.de. In this poster, we introduce the product of DAHITI and the functionality of the DAHITI web service. Furthermore, selected examples of inland water targets are presented in detail. DAHITI provides time series of water level heights of inland water bodies and their formal errors . These time series are available within the period of 1992-2015 and have varying temporal resolutions depending on the data coverage of the investigated water body. The accuracies of the water level time series depend mainly on the extent of the investigated water body and the quality of the altimeter measurements. Hereby, an external validation with in-situ data reveals RMS differences between 5 cm and 40 cm for lakes and 10 cm and 140 cm for rivers, respectively.
Piecewise Trend Approximation: A Ratio-Based Time Series Representation
Jingpei Dan
2013-01-01
Full Text Available A time series representation, piecewise trend approximation (PTA, is proposed to improve efficiency of time series data mining in high dimensional large databases. PTA represents time series in concise form while retaining main trends in original time series; the dimensionality of original data is therefore reduced, and the key features are maintained. Different from the representations that based on original data space, PTA transforms original data space into the feature space of ratio between any two consecutive data points in original time series, of which sign and magnitude indicate changing direction and degree of local trend, respectively. Based on the ratio-based feature space, segmentation is performed such that each two conjoint segments have different trends, and then the piecewise segments are approximated by the ratios between the first and last points within the segments. To validate the proposed PTA, it is compared with classical time series representations PAA and APCA on two classical datasets by applying the commonly used K-NN classification algorithm. For ControlChart dataset, PTA outperforms them by 3.55% and 2.33% higher classification accuracy and 8.94% and 7.07% higher for Mixed-BagShapes dataset, respectively. It is indicated that the proposed PTA is effective for high dimensional time series data mining.
Forecasting Compositional Time Series with Exponential Smoothing Methods
Anne B. Koehler; Ralph D Snyder; J Keith Ord; Adrian Beaumont
2010-01-01
Compositional time series are formed from measurements of proportions that sum to one in each period of time. We might be interested in forecasting the proportion of home loans that have adjustable rates, the proportion of nonagricultural jobs in manufacturing, the proportion of a rock's geochemical composition that is a specific oxide, or the proportion of an election betting market choosing a particular candidate. A problem may involve many related time series of proportions. There could be...
Image-Based Learning Approach Applied to Time Series Forecasting
J. C. Chimal-Eguía; K. Ramírez-Amáro
2012-01-01
In this paper, a new learning approach based on time-series image information is presented. In order to implementthis new learning technique, a novel time-series input data representation is also defined. This input data representation is based on information obtained by image axis division into boxes. The difference between this new input data representation and the classical is that this technique is not time-dependent. This new information is implemented in the new Image-Based Learning A...
Artificial neural networks applied to forecasting time series
Montaño Moreno, Juan José; Palmer Pol, Alfonso; Muñoz Gracia, María del Pilar
2011-01-01
This study offers a description and comparison of the main models of Artificial Neural Networks (ANN) which have proved to be useful in time series forecasting, and also a standard procedure for the practical application of ANN in this type of task. The Multilayer Perceptron (MLP), Radial Base Function (RBF), Generalized Regression Neural Network (GRNN), and Recurrent Neural Network (RNN) models are analyzed. With this aim in mind, we use a time series made up of 244 time points. A comparativ...
Analysis of multidimensional geophysical monitoring time series for earthquake prediction
Lyubushin, A. A.
1999-01-01
A method is presented for detection of synchronous signals in multidimensional time series data. It is based on estimation of eigenvalues of spectral matrices and canonical coherences in moving time windows and extraction of an aggregated signal (a scalar signal, which accumulates in its own variations only those spectral components which are present simultaneously in each scalar time series). It is known that an increase in the collective behavior of the components of some systems and an enl...
Detecting temporal and spatial correlations in pseudoperiodic time series
Zhang, Jie; Luo, Xiaodong; Nakamura, Tomomichi; Sun, Junfeng; Small, Michael
2007-01-01
Recently there has been much attention devoted to exploring the complicated possibly chaotic dynamics in pseudoperiodic time series. Two methods [Zhang , Phys. Rev. E 73, 016216 (2006); Zhang and Small, Phys. Rev. Lett. 96, 238701 (2006)] have been forwarded to reveal the chaotic temporal and spatial correlations, respectively, among the cycles in the time series. Both these methods treat the cycle as the basic unit and design specific statistics that indicate the presence of chaotic dynamics. In this paper, we verify the validity of these statistics to capture the chaotic correlation among cycles by using the surrogate data method. In particular, the statistics computed for the original time series are compared with those from its surrogates. The surrogate data we generate is pseudoperiodic type (PPS), which preserves the inherent periodic components while destroying the subtle nonlinear (chaotic) structure. Since the inherent chaotic correlations among cycles, either spatial or temporal (which are suitably characterized by the proposed statistics), are eliminated through the surrogate generation process, we expect the statistics from the surrogate to take significantly different values than those from the original time series. Hence the ability of the statistics to capture the chaotic correlation in the time series can be validated. Application of this procedure to both chaotic time series and real world data clearly demonstrates the effectiveness of the statistics. We have found clear evidence of chaotic correlations among cycles in human electrocardiogram and vowel time series. Furthermore, we show that this framework is more sensitive to examine the subtle changes in the dynamics of the time series due to the match between PPS surrogate and the statistics adopted. It offers a more reliable tool to reveal the possible correlations among cycles intrinsic to the chaotic nature of the pseudoperiodic time series.
Investigating effects in GNSS station coordinate time series
Haritonova, Diana; Balodis, Janis; Janpaule, Inese
2014-01-01
The vertical and horizontal displacements of the Earth can be measured to a high degree of precision using GNSS. Time series of Latvian GNSS station positions of both the EUPOS®-Riga and LatPos networks have been developed at the Institute of Geodesy and Geoinformation of the University of Latvia (LU GGI). In this study the main focus is made on the noise analysis of the obtained time series and site displacement identification. The results of time series have been analysed and distinctive be...
Analysis of complex time series using refined composite multiscale entropy
Wu, Shuen-De; Wu, Chiu-Wen [Department of Mechatronic Technology, National Taiwan Normal University, Taipei 10610, Taiwan (China); Lin, Shiou-Gwo [Department of Communications, Navigation and Control Engineering, National Taiwan Ocean University, Keelung 20224, Taiwan (China); Lee, Kung-Yen [Department of Engineering Science and Ocean Engineering, National Taiwan University, Taipei 10617, Taiwan (China); Peng, Chung-Kang [College of Health Sciences and Technology, National Central University, Chung-Li 32001, Taiwan (China); Division of Interdisciplinary Medicine and Biotechnology, Beth Israel Deaconess Medical Center/Harvard Medical School, Boston (United States)
2014-04-01
Multiscale entropy (MSE) is an effective algorithm for measuring the complexity of a time series that has been applied in many fields successfully. However, MSE may yield an inaccurate estimation of entropy or induce undefined entropy because the coarse-graining procedure reduces the length of a time series considerably at large scales. Composite multiscale entropy (CMSE) was recently proposed to improve the accuracy of MSE, but it does not resolve undefined entropy. Here we propose a refined composite multiscale entropy (RCMSE) to improve CMSE. For short time series analyses, we demonstrate that RCMSE increases the accuracy of entropy estimation and reduces the probability of inducing undefined entropy.
On the detection of superdiffusive behaviour in time series
Gottwald, Georg A
2016-01-01
We present a new method for detecting superdiffusive behaviour and for determining rates of superdiffusion in time series data. Our method applies equally to stochastic and deterministic time series data and relies on one realisation (ie one sample path) of the process. Linear drift effects are automatically removed without any preprocessing. We show numerical results for time series constructed from i.i.d. $\\alpha$-stable random variables and from deterministic weakly chaotic maps. We compare our method with the standard method of estimating the growth rate of the mean-square displacement as well as the $p$-variation method.
Estimation of connectivity measures in gappy time series
Papadopoulos, G
2015-01-01
A new method is proposed to compute connectivity measures on multivariate time series with gaps. Rather than removing or filling the gaps, the rows of the joint data matrix containing empty entries are removed and the calculations are done on the remainder matrix. The method, called measure adapted gap removal (MAGR), can be applied to any connectivity measure that uses a joint data matrix, such as cross correlation, cross mutual information and transfer entropy. MAGR is favorably compared using these three measures to a number of known gap-filling techniques, as well as the gap closure. The superiority of MAGR is illustrated on time series from synthetic systems and financial time series.
Segmentation of Nonstationary Time Series with Geometric Clustering
Bocharov, Alexei; Thiesson, Bo
2013-01-01
We introduce a non-parametric method for segmentation in regimeswitching time-series models. The approach is based on spectral clustering of target-regressor tuples and derives a switching regression tree, where regime switches are modeled by oblique splits. Such models can be learned efficiently...... from data, where clustering is used to propose one single split candidate at each split level. We use the class of ART time series models to serve as illustration, but because of the non-parametric nature of our segmentation approach, it readily generalizes to a wide range of time-series models that go...
Multivariate time series analysis with R and financial applications
Tsay, Ruey S
2013-01-01
Since the publication of his first book, Analysis of Financial Time Series, Ruey Tsay has become one of the most influential and prominent experts on the topic of time series. Different from the traditional and oftentimes complex approach to multivariate (MV) time series, this sequel book emphasizes structural specification, which results in simplified parsimonious VARMA modeling and, hence, eases comprehension. Through a fundamental balance between theory and applications, the book supplies readers with an accessible approach to financial econometric models and their applications to real-worl
Multi-dimensional sparse time series: feature extraction
Franciosi, Marco
2008-01-01
We show an analysis of multi-dimensional time series via entropy and statistical linguistic techniques. We define three markers encoding the behavior of the series, after it has been translated into a multi-dimensional symbolic sequence. The leading component and the trend of the series with respect to a mobile window analysis result from the entropy analysis and label the dynamical evolution of the series. The diversification formalizes the differentiation in the use of recurrent patterns, from a Zipf law point of view. These markers are the starting point of further analysis such as classification or clustering of large database of multi-dimensional time series, prediction of future behavior and attribution of new data. We also present an application to economic data. We deal with measurements of money investments of some business companies in advertising market for different media sources.
Fast and Flexible Multivariate Time Series Subsequence Search
National Aeronautics and Space Administration — Multivariate Time-Series (MTS) are ubiquitous, and are generated in areas as disparate as sensor recordings in aerospace systems, music and video streams, medical...
Phenotyping of Clinical Time Series with LSTM Recurrent Neural Networks
Lipton, Zachary C.; Kale, David C.; Wetzell, Randall C.
2015-01-01
We present a novel application of LSTM recurrent neural networks to multilabel classification of diagnoses given variable-length time series of clinical measurements. Our method outperforms a strong baseline on a variety of metrics.
A mixed time series model of binomial counts
Khoo, Wooi Chen; Ong, Seng Huat
2015-10-01
Continuous time series modelling has been an active research in the past few decades. However, time series data in terms of correlated counts appear in many situations such as the counts of rainy days and access downloading. Therefore, the study on count data has become popular in time series modelling recently. This article introduces a new mixture model, which is an univariate non-negative stationary time series model with binomial marginal distribution, arising from the combination of the well-known binomial thinning and Pegram's operators. A brief review of important properties will be carried out and the EM algorithm is applied in parameter estimation. A numerical study is presented to show the performance of the model. Finally, a potential real application will be presented to illustrate the advantage of the new mixture model.
Lagrangian Time Series Models for Ocean Surface Drifter Trajectories
Sykulski, Adam M; Lilly, Jonathan M; Danioux, Eric
2016-01-01
This paper proposes stochastic models for the analysis of ocean surface trajectories obtained from freely-drifting satellite-tracked instruments. The proposed time series models are used to summarise large multivariate datasets and infer important physical parameters of inertial oscillations and other ocean processes. Nonstationary time series methods are employed to account for the spatiotemporal variability of each trajectory. Because the datasets are large, we construct computationally efficient methods through the use of frequency-domain modelling and estimation, with the data expressed as complex-valued time series. We detail how practical issues related to sampling and model misspecification may be addressed using semi-parametric techniques for time series, and we demonstrate the effectiveness of our stochastic models through application to both real-world data and to numerical model output.
AFSC/ABL: Naknek sockeye salmon scale time series
National Oceanic and Atmospheric Administration, Department of Commerce — A time series of scale samples (1956 2002) collected from adult sockeye salmon returning to Naknek River were retrieved from the Alaska Department of Fish and Game....
AFSC/ABL: Ugashik sockeye salmon scale time series
National Oceanic and Atmospheric Administration, Department of Commerce — A time series of scale samples (1956 b?? 2002) collected from adult sockeye salmon returning to Ugashik River were retrieved from the Alaska Department of Fish and...
The Ohio economy: using time series characteristics in forecasting
James G. Hoehn; James J. Balazsy
1985-01-01
The premise of this study is that the regional economist can better understand the Ohio economy by studying the properties of important Ohio time series that can be identified and quantified through simple regression methods.
Residual diagnostics for cross-section time series regression models
Baum, Christopher F
2001-01-01
These routines support the diagnosis of groupwise heteroskedasticity and cross-sectional correlation in the context of a regression model fit to pooled cross-section time series (xt) data. Copyright 2001 by Stata Corporation.
Review of English textbooks in time series analysis (in Russian)
Stanislav Anatolyev
2008-01-01
This is a survey of most notable time series econometrics texts written in English. The essay reflects the author's opinion, as well as opinions of econometricians expressed in published book reviews.
A Matlab Code for Univariate Time Series Forecasting
Shapour Mohammadi; Hossein Abbasi- Nejad
2005-01-01
This M-File forecasts univariate time series such as stock prices with a feedforward neural networks. It finds best (minimume RMSE) network automatically and uses early stopping method for solving overfitting problem.
Effect of two dead times in series on coincidence measurements
Funck, E.
1987-01-01
Dead times in series occur in any counting device where detector signals are electronically amplified, selected for pulse height by a discriminator (or pulse-height analyzer) and fed through a dead-time unit producing a dead time, which has very often been designed to establish definite dead time losses. The problem of two dead times in series has been treated by J.W. Muller. No attempt, however, seems to have been made up to now to investigate this problem for the electronics of a coincidence system. In the paper presented here two dead times in series are considered that are found either in one or in both channels of a coincidence system. Correction formulas with experimental evidence are given which allow the deviations from results, which were calculated by taking only one dead time per channel into account, to be estimated.
Neural Networks, Game Theory and Time Series Generation
Metzler, Richard
2002-01-01
This dissertation highlights connections between the fields of neural networks, game theory and time series generation. The concept of antipredictability is explained, and the properties of time series that are antipredictable for several prototypical prediction algorithms (neural networks, Boolean funtions etc.) are studied. The Minority Game provides a framework in which antipredictability arises naturally. Several variations of the MG are introduced and compared, including extensions to mo...
The use of synthetic input sequences in time series modeling
Oliveira, Dair Jose de [Programa de Pos-Graduacao em Engenharia Eletrica, Universidade Federal de Minas Gerais, Av. Antonio Carlos 6627, 31.270-901 Belo Horizonte, MG (Brazil); Letellier, Christophe [CORIA/CNRS UMR 6614, Universite et INSA de Rouen, Av. de l' Universite, BP 12, F-76801 Saint-Etienne du Rouvray cedex (France); Gomes, Murilo E.D. [Programa de Pos-Graduacao em Engenharia Eletrica, Universidade Federal de Minas Gerais, Av. Antonio Carlos 6627, 31.270-901 Belo Horizonte, MG (Brazil); Aguirre, Luis A. [Programa de Pos-Graduacao em Engenharia Eletrica, Universidade Federal de Minas Gerais, Av. Antonio Carlos 6627, 31.270-901 Belo Horizonte, MG (Brazil)], E-mail: aguirre@cpdee.ufmg.br
2008-08-04
In many situations time series models obtained from noise-like data settle to trivial solutions under iteration. This Letter proposes a way of producing a synthetic (dummy) input, that is included to prevent the model from settling down to a trivial solution, while maintaining features of the original signal. Simulated benchmark models and a real time series of RR intervals from an ECG are used to illustrate the procedure.
Stacked Heterogeneous Neural Networks for Time Series Forecasting
Florin Leon
2010-01-01
Full Text Available A hybrid model for time series forecasting is proposed. It is a stacked neural network, containing one normal multilayer perceptron with bipolar sigmoid activation functions, and the other with an exponential activation function in the output layer. As shown by the case studies, the proposed stacked hybrid neural model performs well on a variety of benchmark time series. The combination of weights of the two stack components that leads to optimal performance is also studied.
Model-Coupled Autoencoder for Time Series Visualisation
Gianniotis, Nikolaos; Kügler, Sven D.; Tiňo, Peter; Polsterer, Kai L.
2016-01-01
We present an approach for the visualisation of a set of time series that combines an echo state network with an autoencoder. For each time series in the dataset we train an echo state network, using a common and fixed reservoir of hidden neurons, and use the optimised readout weights as the new representation. Dimensionality reduction is then performed via an autoencoder on the readout weight representations. The crux of the work is to equip the autoencoder with a loss function that correctl...
On the prediction of stationary functional time series
Aue, A.; Norinho, DD; Hörmann, S
2012-01-01
© 2015, American Statistical Association. This article addresses the prediction of stationary functional time series. Existing contributions to this problem have largely focused on the special case of first-order functional autoregressive processes because of their technical tractability and the current lack of advanced functional time series methodology. It is shown here how standard multivariate prediction techniques can be used in this context. The connection between functional and multiva...
Trimmed Granger causality between two groups of time series
Hung, Ying-Chao; Tseng, Neng-Fang; Balakrishnan, Narayanaswamy
2014-01-01
The identification of causal effects between two groups of time series has been an important topic in a wide range of applications such as economics, engineering, medicine, neuroscience, and biology. In this paper, a simplified causal relationship (called trimmed Granger causality) based on the context of Granger causality and vector autoregressive (VAR) model is introduced. The idea is to characterize a subset of “important variables” for both groups of time series so that the underlying cau...
Time series analysis of age related cataract hospitalizations and phacoemulsification
2006-01-01
Background Cataract surgery remains a commonly performed elective surgical procedure in the aging and the elderly. The purpose of this study was to utilize time series methodology to determine the temporal and seasonal variations and the strength of the seasonality in age-related (senile) cataract hospitalizations and phacoemulsification surgeries. Methods A retrospective, cross-sectional time series analysis was used to assess the presence and strength of seasonal and temporal patterns of ag...
Gaussian Processes for Local Polynomial Forecasting of Time Series
Fendick, Kerry
2016-01-01
The samples of a signal obscured by noise constitute an example of a time series frequently encountered in applications. We consider here the feasibility of accurately forecasting the signals of multiple such time series considering jointly when the number of historic samples is inadequate for accurately forecasting the signal of each considered in isolation. We develop a new forecasting methodology based on Gaussian process regression that is successful in doing so in examples for which the ...
Time Series Estimates of the Italian Consumer Confidence Indicator
Paradiso, Antonio; Rao, B. Bhaskara; Margani, Patrizia
2011-01-01
This work shows that Italian consumer confidence indicator (CCI) is non-stationary and, therefore, can be estimated with the time series methods. It is found that a long-run relationship exists between CCI, short-term interest rate, industrial production index and the difference between perceived and measured inflation. The use of time series methods to estimate CCI for Italy is a novelty in the literature.
From Local to Global Analysis of Music Time Series
Ligges, Uwe; Weihs, Claus
2004-01-01
Local and more and more global musical structure is analyzed from audio time series by time-series-event analysis with the aim of automatic sheet music production and comparison of singers. Note events are determined and classified based on local spectra, and rules of bar events are identified based on accentuation events related to local energy. In order to compare the performances of different singers global summary measures are defined characterizing the overall performance.
Multiple Time Series Ising Model for Financial Market Simulations
In this paper we propose an Ising model which simulates multiple financial time series. Our model introduces the interaction which couples to spins of other systems. Simulations from our model show that time series exhibit the volatility clustering that is often observed in the real financial markets. Furthermore we also find non-zero cross correlations between the volatilities from our model. Thus our model can simulate stock markets where volatilities of stocks are mutually correlated
Geomechanical time series and its singularity spectrum analysis
Lyubushin, Alexei A.; Kaláb, Zdeněk; Lednická, Markéta
2012-01-01
Roč. 47, č. 1 (2012), s. 69-77. ISSN 1217-8977 R&D Projects: GA ČR GA105/09/0089 Institutional research plan: CEZ:AV0Z30860518 Keywords : geomechanical time series * singularity spectrum * time series segmentation * laser distance meter Subject RIV: DC - Siesmology, Volcanology, Earth Structure Impact factor: 0.347, year: 2012 http://www.akademiai.com/content/88v4027758382225/fulltext.pdf
Time Series Classification by Class-Specific Mahalanobis Distance Measures
Prekopcsák, Zoltán; Lemire, Daniel
2010-01-01
To classify time series by nearest neighbors, we need to specify or learn one or several distance measures. We consider variations of the Mahalanobis distance measures which rely on the inverse covariance matrix of the data. Unfortunately --- for time series data --- the covariance matrix has often low rank. To alleviate this problem we can either use a pseudoinverse, covariance shrinking or limit the matrix to its diagonal. We review these alternatives and benchmark them against competitive ...
Automated Feature Design for Time Series Classification by Genetic Programming
Harvey, Dustin Yewell
2014-01-01
Time series classification (TSC) methods discover and exploit patterns in time series and other one-dimensional signals. Although many accurate, robust classifiers exist for multivariate feature sets, general approaches are needed to extend machine learning techniques to make use of signal inputs. Numerous applications of TSC can be found in structural engineering, especially in the areas of structural health monitoring and non-destructive evaluation. Additionally, the fields of process contr...
Application of a Local Polynomial Approximation Chaotic Time Series Prediction
Orzeszko, Witold
2004-01-01
Chaos theory has become a new approach to financial processes analysis. Due to complicated dynamics, chaotic time series seem to be random and, in consequence, unpredictable. In fact, unlike truly random processes, chaotic dynamics can be forecasted very precisely in a short run. In this paper, a local polynomial approximation is presented. Its efficiency, as a method of building short-term predictors of chaotic time series, has been examined. The presented method has been applied to forecast...
Adaptive Fourier Analysis For Unequally-Spaced Time Series Data
Liang, Hong
2002-01-01
Adaptive Fourier Analysis For Unequally-Spaced Time Series Data by Hong Liang Robert V. Foutz, Chairman Statistics (ABSTRACT) Fourier analysis, Walsh-Fourier analysis, and wavelet analysis have often been used in time series analysis. Fourier analysis can be used to detect periodic components that have sinusoidal shape; however, it might be misleading when the periodic components are not sinusoidal. Walsh-Fourier analysis is suitable for revealing the rectangular ...
Locally adaptive factor processes for multivariate time series
Durante, Daniele; Scarpa, Bruno; Dunson, David B
2012-01-01
In modeling multivariate time series, it is important to allow time-varying smoothness in the mean and covariance process. In particular, there may be certain time intervals exhibiting rapid changes and others in which changes are slow. If such time-varying smoothness is not accounted for, one can obtain misleading inferences and predictions, with over-smoothing across erratic time intervals and under-smoothing across times exhibiting slow variation. This can lead to mis-calibration of predic...
Time series modelling and forecasting of Sarawak black pepper price
Liew, Venus Khim-Sen; Shitan, Mahendran; Hussain, Huzaimi
2000-01-01
Pepper is an important agriculture commodity especially for the state of Sarawak. It is important to forecast its price, as this could help the policy makers in coming up with production and marketing plan to improve the Sarawak’s economy as well as the farmers’welfare. In this paper, we take up time series modelling and forecasting of the Sarawak black pepper price. Our empirical results show that Autoregressive Moving Average (ARMA) time series models fit the price series well and they have...
Time Series Analysis of Insar Data: Methods and Trends
Osmanoglu, Batuhan; Sunar, Filiz; Wdowinski, Shimon; Cano-Cabral, Enrique
2015-01-01
Time series analysis of InSAR data has emerged as an important tool for monitoring and measuring the displacement of the Earth's surface. Changes in the Earth's surface can result from a wide range of phenomena such as earthquakes, volcanoes, landslides, variations in ground water levels, and changes in wetland water levels. Time series analysis is applied to interferometric phase measurements, which wrap around when the observed motion is larger than one-half of the radar wavelength. Thus, the spatio-temporal ''unwrapping" of phase observations is necessary to obtain physically meaningful results. Several different algorithms have been developed for time series analysis of InSAR data to solve for this ambiguity. These algorithms may employ different models for time series analysis, but they all generate a first-order deformation rate, which can be compared to each other. However, there is no single algorithm that can provide optimal results in all cases. Since time series analyses of InSAR data are used in a variety of applications with different characteristics, each algorithm possesses inherently unique strengths and weaknesses. In this review article, following a brief overview of InSAR technology, we discuss several algorithms developed for time series analysis of InSAR data using an example set of results for measuring subsidence rates in Mexico City.
Time-varying parameter auto-regressive models for autocovariance nonstationary time series
无
2009-01-01
In this paper, autocovariance nonstationary time series is clearly defined on a family of time series. We propose three types of TVPAR (time-varying parameter auto-regressive) models: the full order TVPAR model, the time-unvarying order TVPAR model and the time-varying order TV-PAR model for autocovariance nonstationary time series. Related minimum AIC (Akaike information criterion) estimations are carried out.
Time-varying parameter auto-regressive models for autocovariance nonstationary time series
FEI WanChun; BAI Lun
2009-01-01
In this paper,autocovariance nonstationary time series is clearly defined on a family of time series.We propose three types of TVPAR (time-varying parameter auto-regressive) models:the full order TVPAR model,the time-unvarying order TVPAR model and the time-varying order TVPAR model for autocovariance nonstationary time series.Related minimum AIC (Akaike information criterion) estimations are carried out.
Comparison of New and Old Sunspot Number Time Series
Cliver, Edward W.; Clette, Frédéric; Lefévre, Laure; Svalgaard, Leif
2016-05-01
As a result of the Sunspot Number Workshops, five new sunspot series have recently been proposed: a revision of the original Wolf or international sunspot number (Lockwood et al., 2014), a backbone-based group sunspot number (Svalgaard and Schatten, 2016), a revised group number series that employs active day fractions (Usoskin et al., 2016), a provisional group sunspot number series (Cliver and Ling, 2016) that removes flaws in the normalization scheme for the original group sunspot number (Hoyt and Schatten,1998), and a revised Wolf or international number (termed SN) published on the SILSO website as a replacement for the original Wolf number (Clette and Lefèvre, 2016; thttp://www.sidc.be/silso/datafiles). Despite quite different construction methods, the five new series agree reasonably well after about 1900. From 1750 to ~1875, however, the Lockwood et al. and Usoskin et al. time series are lower than the other three series. Analysis of the Hoyt and Schatten normalization factors used to scale secondary observers to their Royal Greenwich Observatory primary observer reveals a significant inhomogeneity spanning the divergence in ~1885 of the group number from the original Wolf number. In general, a correction factor time series, obtained by dividing an annual group count series by the corresponding yearly averages of raw group counts for all observers, can be used to assess the reliability of new sunspot number reconstructions.
Symplectic geometry spectrum regression for prediction of noisy time series
Xie, Hong-Bo; Dokos, Socrates; Sivakumar, Bellie; Mengersen, Kerrie
2016-05-01
We present the symplectic geometry spectrum regression (SGSR) technique as well as a regularized method based on SGSR for prediction of nonlinear time series. The main tool of analysis is the symplectic geometry spectrum analysis, which decomposes a time series into the sum of a small number of independent and interpretable components. The key to successful regularization is to damp higher order symplectic geometry spectrum components. The effectiveness of SGSR and its superiority over local approximation using ordinary least squares are demonstrated through prediction of two noisy synthetic chaotic time series (Lorenz and Rössler series), and then tested for prediction of three real-world data sets (Mississippi River flow data and electromyographic and mechanomyographic signal recorded from human body).
Similarity estimators for irregular and age uncertain time series
K. Rehfeld
2013-09-01
Full Text Available Paleoclimate time series are often irregularly sampled and age uncertain, which is an important technical challenge to overcome for successful reconstruction of past climate variability and dynamics. Visual comparison and interpolation-based linear correlation approaches have been used to infer dependencies from such proxy time series. While the first is subjective, not measurable and not suitable for the comparison of many datasets at a time, the latter introduces interpolation bias, and both face difficulties if the underlying dependencies are nonlinear. In this paper we investigate similarity estimators that could be suitable for the quantitative investigation of dependencies in irregular and age uncertain time series. We compare the Gaussian-kernel based cross correlation (gXCF, Rehfeld et al., 2011 and mutual information (gMI, Rehfeld et al., 2013 against their interpolation-based counterparts and the new event synchronization function (ESF. We test the efficiency of the methods in estimating coupling strength and coupling lag numerically, using ensembles of synthetic stalagmites with short, autocorrelated, linear and nonlinearly coupled proxy time series, and in the application to real stalagmite time series. In the linear test case coupling strength increases are identified consistently for all estimators, while in the nonlinear test case the correlation-based approaches fail. The lag at which the time series are coupled is identified correctly as the maximum of the similarity functions in around 60–55% (in the linear case to 53–42% (for the nonlinear processes of the cases when the dating of the synthetic stalagmite is perfectly precise. If the age uncertainty increases beyond 5% of the time series length, however, the true coupling lag is not identified more often than the others for which the similarity function was estimated. Age uncertainty contributes up to half of the uncertainty in the similarity estimation process. Time
Detection of flood events in hydrological discharge time series
Seibert, S. P.; Ehret, U.
2012-04-01
The shortcomings of mean-squared-error (MSE) based distance metrics are well known (Beran 1999, Schaeffli & Gupta 2007) and the development of novel distance metrics (Pappenberger & Beven 2004, Ehret & Zehe 2011) and multi-criteria-approaches enjoy increasing popularity (Reusser 2009, Gupta et al. 2009). Nevertheless, the hydrological community still lacks metrics which identify and thus, allow signature based evaluations of hydrological discharge time series. Signature based information/evaluations are required wherever specific time series features, such as flood events, are of special concern. Calculation of event based runoff coefficients or precise knowledge on flood event characteristics (like onset or duration of rising limp or the volume of falling limp, etc.) are possible applications. The same applies for flood forecasting/simulation models. Directly comparing simulated and observed flood event features may reveal thorough insights into model dynamics. Compared to continuous space-and-time-aggregated distance metrics, event based evaluations may provide answers like the distributions of event characteristics or the percentage of the events which were actually reproduced by a hydrological model. It also may help to provide information on the simulation accuracy of small, medium and/or large events in terms of timing and magnitude. However, the number of approaches which expose time series features is small and their usage is limited to very specific questions (Merz & Blöschl 2009, Norbiato et al. 2009). We believe this is due to the following reasons: i) a generally accepted definition of the signature of interest is missing or difficult to obtain (in our case: what makes a flood event a flood event?) and/or ii) it is difficult to translate such a definition into a equation or (graphical) procedure which exposes the feature of interest in the discharge time series. We reviewed approaches which detect event starts and/or ends in hydrological discharge time
Wavelet matrix transform for time-series similarity measurement
HU Zhi-kun; XU Fei; GUI Wei-hua; YANG Chun-hua
2009-01-01
A time-series similarity measurement method based on wavelet and matrix transform was proposed, and its anti-noise ability, sensitivity and accuracy were discussed. The time-series sequences were compressed into wavelet subspace, and sample feature vector and orthogonal basics of sample time-series sequences were obtained by K-L transform. Then the inner product transform was carried out to project analyzed time-series sequence into orthogonal basics to gain analyzed feature vectors. The similarity was calculated between sample feature vector and analyzed feature vector by the Euclid distance. Taking fault wave of power electronic devices for example, the experimental results show that the proposed method has low dimension of feature vector, the anti-noise ability of proposed method is 30 times as large as that of plain wavelet method, the sensitivity of proposed method is 1/3 as large as that of plain wavelet method, and the accuracy of proposed method is higher than that of the wavelet singular value decomposition method. The proposed method can be applied in similarity matching and indexing for lager time series databases.
Discovering shared and individual latent structure in multiple time series
Saria, Suchi; Penn, Anna
2010-01-01
This paper proposes a nonparametric Bayesian method for exploratory data analysis and feature construction in continuous time series. Our method focuses on understanding shared features in a set of time series that exhibit significant individual variability. Our method builds on the framework of latent Diricihlet allocation (LDA) and its extension to hierarchical Dirichlet processes, which allows us to characterize each series as switching between latent ``topics'', where each topic is characterized as a distribution over ``words'' that specify the series dynamics. However, unlike standard applications of LDA, we discover the words as we learn the model. We apply this model to the task of tracking the physiological signals of premature infants; our model obtains clinically significant insights as well as useful features for supervised learning tasks.
Forecasting Compositional Time Series: A State Space Approach
Ralph D Snyder; J. Keith Ord; Anne B. Koehler; Keith R. McLaren; Adrian Beaumont
2015-01-01
A method is proposed for forecasting composite time series such as the market shares for multiple brands. Its novel feature is that it relies on multi-series adaptations of exponential smoothing combined with the log-ratio transformation for the conversion of proportions onto the real line. It is designed to produce forecasts that are both non-negative and sum to one; are invariant to the choice of the base series in the log-ratio transformation; recognized and exploit features such as serial...
Arbitrage, market definition and monitoring a time series approach
Burke, S.; Hunter, J
2012-01-01
This article considers the application to regional price data of time series methods to test stationarity, multivariate cointegration and exogeneity. The discovery of stationary price differentials in a bivariate setting implies that the series are rendered stationary by capturing a common trend and we observe through this mechanism long-run arbitrage. This is indicative of a broader market definition and efficiency. The problem is considered in relation to more than 700 weekly data points on...
Detecting structural breaks in time series via genetic algorithms
Doerr, Benjamin; Fischer, Paul; Hilbert, Astrid;
2016-01-01
Detecting structural breaks is an essential task for the statistical analysis of time series, for example, for fitting parametric models to it. In short, structural breaks are points in time at which the behaviour of the time series substantially changes. Typically, no solid background knowledge of...... crossover and mutation operations for this problem, we conduct extensive experiments to determine good choices for the parameters and operators of the genetic algorithm. One surprising observation is that use of uniform and one-point crossover together gave significantly better results than using either...... crossover operator alone. Moreover, we present a specific fitness function which exploits the sparse structure of the break points and which can be evaluated particularly efficiently. The experiments on artificial and real-world time series show that the resulting algorithm detects break points with high...
Weighted statistical parameters for irregularly sampled time series
Rimoldini, Lorenzo
2014-01-01
Unevenly spaced time series are common in astronomy because of the day-night cycle, weather conditions, dependence on the source position in the sky, allocated telescope time, corrupt measurements, for example, or be inherent to the scanning law of satellites like Hipparcos and the forthcoming Gaia. This paper aims at improving the accuracy of common statistical parameters for the characterization of irregularly sampled signals. The uneven representation of time series, often including clumps of measurements and gaps with no data, can severely disrupt the values of estimators. A weighting scheme adapting to the sampling density and noise level of the signal is formulated. Its application to time series from the Hipparcos periodic catalogue led to significant improvements in the overall accuracy and precision of the estimators with respect to the unweighted counterparts and those weighted by inverse-squared uncertainties. Automated classification procedures employing statistical parameters weighted by the sugg...
Rules extraction in short memory time series using genetic algorithms
Fong, L. Y.; Szeto, K. Y.
2001-04-01
Data mining is performed using genetic algorithm on artificially generated time series data with short memory. The extraction of rules from a training set and the subsequent testing of these rules provide a basis for the predictions on the test set. The artificial time series are generated using the inverse whitening transformation, and the correlation function has an exponential form with given time constant indicative of short memory. A vector quantization technique is employed to classify the daily rate of return of this artificial time series into four categories. A simple genetic algorithm based on a fixed format of rules is introduced to do the forecasting. Comparing to the benchmark tests with random walk and random guess, genetic algorithms yield substantially better prediction rates, between 50% to 60%. This is an improvement compared with the 47% for random walk prediction and 25% for random guessing method.
A refined fuzzy time series model for stock market forecasting
Jilani, Tahseen Ahmed; Burney, Syed Muhammad Aqil
2008-05-01
Time series models have been used to make predictions of stock prices, academic enrollments, weather, road accident casualties, etc. In this paper we present a simple time-variant fuzzy time series forecasting method. The proposed method uses heuristic approach to define frequency-density-based partitions of the universe of discourse. We have proposed a fuzzy metric to use the frequency-density-based partitioning. The proposed fuzzy metric also uses a trend predictor to calculate the forecast. The new method is applied for forecasting TAIEX and enrollments’ forecasting of the University of Alabama. It is shown that the proposed method work with higher accuracy as compared to other fuzzy time series methods developed for forecasting TAIEX and enrollments of the University of Alabama.
Generalized Dynamic Factor Models for Mixed-Measurement Time Series.
Cui, Kai; Dunson, David B
2014-02-12
In this article, we propose generalized Bayesian dynamic factor models for jointly modeling mixed-measurement time series. The framework allows mixed-scale measurements associated with each time series, with different measurements having different distributions in the exponential family conditionally on time-varying latent factor(s). Efficient Bayesian computational algorithms are developed for posterior inference on both the latent factors and model parameters, based on a Metropolis Hastings algorithm with adaptive proposals. The algorithm relies on a Greedy Density Kernel Approximation (GDKA) and parameter expansion with latent factor normalization. We tested the framework and algorithms in simulated studies and applied them to the analysis of intertwined credit and recovery risk for Moody's rated firms from 1982-2008, illustrating the importance of jointly modeling mixed-measurement time series. The article has supplemental materials available online. PMID:24791133
Minimum entropy density method for the time series analysis
Lee, Jeong Won; Park, Joongwoo Brian; Jo, Hang-Hyun; Yang, Jae-Suk; Moon, Hie-Tae
2009-01-01
The entropy density is an intuitive and powerful concept to study the complicated nonlinear processes derived from physical systems. We develop the minimum entropy density method (MEDM) to detect the structure scale of a given time series, which is defined as the scale in which the uncertainty is minimized, hence the pattern is revealed most. The MEDM is applied to the financial time series of Standard and Poor’s 500 index from February 1983 to April 2006. Then the temporal behavior of structure scale is obtained and analyzed in relation to the information delivery time and efficient market hypothesis.
Multiscale entropy analysis of complex physiologic time series.
Costa, Madalena; Goldberger, Ary L; Peng, C-K
2002-08-01
There has been considerable interest in quantifying the complexity of physiologic time series, such as heart rate. However, traditional algorithms indicate higher complexity for certain pathologic processes associated with random outputs than for healthy dynamics exhibiting long-range correlations. This paradox may be due to the fact that conventional algorithms fail to account for the multiple time scales inherent in healthy physiologic dynamics. We introduce a method to calculate multiscale entropy (MSE) for complex time series. We find that MSE robustly separates healthy and pathologic groups and consistently yields higher values for simulated long-range correlated noise compared to uncorrelated noise. PMID:12190613
Wavelet analysis for non-stationary, nonlinear time series
Schulte, Justin A.
2016-08-01
Methods for detecting and quantifying nonlinearities in nonstationary time series are introduced and developed. In particular, higher-order wavelet analysis was applied to an ideal time series and the quasi-biennial oscillation (QBO) time series. Multiple-testing problems inherent in wavelet analysis were addressed by controlling the false discovery rate. A new local autobicoherence spectrum facilitated the detection of local nonlinearities and the quantification of cycle geometry. The local autobicoherence spectrum of the QBO time series showed that the QBO time series contained a mode with a period of 28 months that was phase coupled to a harmonic with a period of 14 months. An additional nonlinearly interacting triad was found among modes with periods of 10, 16 and 26 months. Local biphase spectra determined that the nonlinear interactions were not quadratic and that the effect of the nonlinearities was to produce non-smoothly varying oscillations. The oscillations were found to be skewed so that negative QBO regimes were preferred, and also asymmetric in the sense that phase transitions between the easterly and westerly phases occurred more rapidly than those from westerly to easterly regimes.
The Photoplethismographic Signal Processed with Nonlinear Time Series Analysis Tools
Finger photoplethismography (PPG) signals were submitted to nonlinear time series analysis. The applied analytical techniques were: (i) High degree polynomial fitting for baseline estimation; (ii) FFT analysis for estimating power spectra; (iii) fractal dimension estimation via the Higuchi's time-domain method, and (iv) kernel nonparametric estimation for reconstructing noise free-attractors and also for estimating signal's stochastic components
Sparse time series chain graphical models for reconstructing genetic networks
Abegaz, Fentaw; Wit, Ernst
2013-01-01
We propose a sparse high-dimensional time series chain graphical model for reconstructing genetic networks from gene expression data parametrized by a precision matrix and autoregressive coefficient matrix. We consider the time steps as blocks or chains. The proposed approach explores patterns of co
AMP: a new time-frequency feature extraction method for intermittent time-series data
Barrack, Duncan; Goulding, James; Hopcraft, Keith; Preston, Simon; Smith, Gavin
2015-01-01
The characterisation of time-series data via their most salient features is extremely important in a range of machine learning task, not least of all with regards to classification and clustering. While there exist many feature extraction techniques suitable for non-intermittent time-series data, these approaches are not always appropriate for intermittent time-series data, where intermittency is characterized by constant values for large periods of time punctuated by sharp and transient incr...
Real Time Clustering of Time Series Using Triangular Potentials
Aldo Pacchiano
2015-01-01
Full Text Available Motivated by the problem of computing investment portfolio weightin gs we investigate various methods of clustering as alternatives to traditional mean-v ariance approaches. Such methods can have significant benefits from a practical point of view since they remove the need to invert a sample covariance matrix, which can suffer from estimation error and will almost certainly be non-stationary. The general idea is to find groups of assets w hich share similar return characteristics over time and treat each group as a singl e composite asset. We then apply inverse volatility weightings to these new composite assets. In the course of our investigation we devise a method of clustering based on triangular potentials and we present as sociated theoretical results as well as various examples based on synthetic data.
Time Series Outlier Detection Based on Sliding Window Prediction
Yufeng Yu
2014-01-01
Full Text Available In order to detect outliers in hydrological time series data for improving data quality and decision-making quality related to design, operation, and management of water resources, this research develops a time series outlier detection method for hydrologic data that can be used to identify data that deviate from historical patterns. The method first built a forecasting model on the history data and then used it to predict future values. Anomalies are assumed to take place if the observed values fall outside a given prediction confidence interval (PCI, which can be calculated by the predicted value and confidence coefficient. The use of PCI as threshold is mainly on the fact that it considers the uncertainty in the data series parameters in the forecasting model to address the suitable threshold selection problem. The method performs fast, incremental evaluation of data as it becomes available, scales to large quantities of data, and requires no preclassification of anomalies. Experiments with different hydrologic real-world time series showed that the proposed methods are fast and correctly identify abnormal data and can be used for hydrologic time series analysis.
Entropy measure of stepwise component in GPS time series
Lyubushin, A. A.; Yakovlev, P. V.
2016-01-01
A new method for estimating the stepwise component in the time series is suggested. The method is based on the application of a pseudo-derivative. The advantage of this method lies in the simplicity of its practical implementation compared to the more common methods for identifying the peculiarities in the time series against the noise. The need for automatic detection of the jumps in the noised signal and for introducing a quantitative measure of a stepwise behavior of the signal arises in the problems of the GPS time series analysis. The interest in the jumps in the mean level of the GPS signal is associated with the fact that they may reflect the typical earthquakes or the so-called silent earthquakes. In this paper, we offer the criteria for quantifying the degree of the stepwise behavior of the noised time series. These criteria are based on calculating the entropy for the auxiliary series of averaged stepwise approximations, which are constructed with the use of pseudo-derivatives.
Multi-Scale Dissemination of Time Series Data
Guo, Qingsong; Zhou, Yongluan; Su, Li
2013-01-01
, which is an abstract indicator for both the physical limits and the amount of data that the subscriber would like to handle. To handle this problem, we propose a system framework for multi-scale time series data dissemination that employs a typical tree-based dissemination network and existing time......In this paper, we consider the problem of continuous dissemination of time series data, such as sensor measurements, to a large number of subscribers. These subscribers fall into multiple subscription levels, where each subscription level is specified by the bandwidth constraint of a subscriber......-series compression models. Due to the bandwidth limits regarding to potentially sheer speed of data, it is inevitable to compress and re-compress data along the dissemination paths according to the subscription level of each node. Compression would caused the accuracy loss of data, thus we devise several algorithms...
Grammar-based feature generation for time-series prediction
De Silva, Anthony Mihirana
2015-01-01
This book proposes a novel approach for time-series prediction using machine learning techniques with automatic feature generation. Application of machine learning techniques to predict time-series continues to attract considerable attention due to the difficulty of the prediction problems compounded by the non-linear and non-stationary nature of the real world time-series. The performance of machine learning techniques, among other things, depends on suitable engineering of features. This book proposes a systematic way for generating suitable features using context-free grammar. A number of feature selection criteria are investigated and a hybrid feature generation and selection algorithm using grammatical evolution is proposed. The book contains graphical illustrations to explain the feature generation process. The proposed approaches are demonstrated by predicting the closing price of major stock market indices, peak electricity load and net hourly foreign exchange client trade volume. The proposed method ...
Increment Entropy as a Measure of Complexity for Time Series
Xiaofeng Liu
2016-01-01
Full Text Available Entropy has been a common index to quantify the complexity of time series in a variety of fields. Here, we introduce an increment entropy to measure the complexity of time series in which each increment is mapped onto a word of two letters, one corresponding to the sign and the other corresponding to the magnitude. Increment entropy (IncrEn is defined as the Shannon entropy of the words. Simulations on synthetic data and tests on epileptic electroencephalogram (EEG signals demonstrate its ability of detecting abrupt changes, regardless of the energetic (e.g., spikes or bursts or structural changes. The computation of IncrEn does not make any assumption on time series, and it can be applicable to arbitrary real-world data.
Detection of "noisy" chaos in a time series
Chon, K H; Kanters, J K; Cohen, R J; Holstein-Rathlou, N H
Time series from biological system often displays fluctuations in the measured variables. Much effort has been directed at determining whether this variability reflects deterministic chaos, or whether it is merely "noise". The output from most biological systems is probably the result of both the...... internal dynamics of the systems, and the input to the system from the surroundings. This implies that the system should be viewed as a mixed system with both stochastic and deterministic components. We present a method that appears to be useful in deciding whether determinism is present in a time series......, and if this determinism has chaotic attributes. The method relies on fitting a nonlinear autoregressive model to the time series followed by an estimation of the characteristic exponents of the model over the observed probability distribution of states for the system. The method is tested by computer...
Increment entropy as a measure of complexity for time series
Liu, Xiaofeng; Xu, Ning; Xue, Jianru
2015-01-01
Entropy has been a common index to quantify the complexity of time series in a variety of fields. Here, we introduce increment entropy to measure the complexity of time series in which each increment is mapped into a word of two letters, one letter corresponding to direction and the other corresponding to magnitude. The Shannon entropy of the words is termed as increment entropy (IncrEn). Simulations on synthetic data and tests on epileptic EEG signals have demonstrated its ability of detecting the abrupt change, regardless of energetic (e.g. spikes or bursts) or structural changes. The computation of IncrEn does not make any assumption on time series and it can be applicable to arbitrary real-world data.
Feature-preserving interpolation and filtering of environmental time series
Mariethoz, Gregoire; Jougnot, Damien; Rezaee, Hassan
2015-01-01
We propose a method for filling gaps and removing interferences in time series for applications involving continuous monitoring of environmental variables. The approach is non-parametric and based on an iterative pattern-matching between the affected and the valid parts of the time series. It considers several variables jointly in the pattern matching process and allows preserving linear or non-linear dependences between variables. The uncertainty in the reconstructed time series is quantified through multiple realizations. The method is tested on self-potential data that are affected by strong interferences as well as data gaps, and the results show that our approach allows reproducing the spectral features of the original signal. Even in the presence of intense signal perturbations, it significantly improves the signal and corrects bias introduced by asymmetrical interferences. Potential applications are wide-ranging, including geophysics, meteorology and hydrology.
Learning of time series through neuron-to-neuron instruction
A model neuron with delayline feedback connections can learn a time series generated by another model neuron. It has been known that some student neurons that have completed such learning under the instruction of a teacher's quasi-periodic sequence mimic the teacher's time series over a long interval, even after instruction has ceased. We found that in addition to such faithful students, there are unfaithful students whose time series eventually diverge exponentially from that of the teacher. In order to understand the circumstances that allow for such a variety of students, the orbit dimension was estimated numerically. The quasi-periodic orbits in question were found to be confined in spaces with dimensions significantly smaller than that of the full phase space
Asymptotics for Nonlinear Transformations of Fractionally Integrated Time Series
无
2007-01-01
The asymptotic theory for nonlinear transformations of fractionally integrated time series is developed. By the use of fractional Occupation Times Formula, various nonlinear functions of fractionally integrated series such as ARFIMA time series are studied, and the asymptotic distributions of the sample moments of such functions are obtained and analyzed. The transformations considered in this paper includes a variety of functions such as regular functions, integrable functions and asymptotically homogeneous functions that are often used in practical nonlinear econometric analysis. It is shown that the asymptotic theory of nonlinear transformations of original and normalized fractionally integrated processes is different from that of fractionally integrated processes, but is similar to the asymptotic theory of nonlinear transformations of integrated processes.
Reconstruction of tritium time series in precipitation of Beijing
Human nuclear activities, especially intensive nuclear tests during the 1960s in the world, caused atmospheric tritium concentration increasing significantly, which provided convenient condition for global water cycle research, especially for tracer dating research of groundwater. However, due to the layout of monitoring sites and other reasons, most parts of the world are lack of monitoring data of tritium concentration in precipitation, which brought difficulties in determining the input function which is essential for groundwater tracer dating technique. Based on the analysis of principles and applicability of present reconstruction methods of tritium time series, the tritium time series in precipitation in Beijing during 1953-2002 was reconstructed using combined methods, including Guanbingjun method, trend surface analysis method, trigonometric interpolation method and correlation method. Furthermore, the best method and the best time series curve were elected through comparison of results of different methods. (authors)
Time series analysis in astronomy: Limits and potentialities
Vio, R.; Kristensen, N.R.; Madsen, Henrik; Wamsteker, W.
2005-01-01
priori physical model there are not many possibilities to obtain interpretable results. For this reason, the practice to develop more and more sophisticated statistical methods of time series analysis is not productive. Only techniques of data analysis developed in a specific physical context can be......In this paper we consider the problem of the limits concerning the physical information that can be extracted from the analysis of one or more time series ( light curves) typical of astrophysical objects. On the basis of theoretical considerations and numerical simulations, we show that with no a...... expected to provide useful results. The field of stochastic dynamics appears to be an interesting framework for such an approach. In particular, it is shown that modelling the experimental time series by means of the stochastic differential equations (SDE) represents a valuable tool of analysis. For...