Online Segmentation of Time Series Based on Polynomial Least-Squares Approximations

The paper presents SwiftSeg, a novel technique for online time series segmentation and piecewise polynomial representation. The segmentation approach is based on a least-squares approximation of time series in sliding and/or growing time windows utilizing a basis of orthogonal polynomials. This allows the definition of fast update steps for the approximating polynomial, where the computational effort depends only on the degree of the approximating polynomial and not on the length of the time window. The coefficients of the orthogonal expansion of the approximating polynomial-obtained by means of the update steps-can be interpreted as optimal (in the least-squares sense) estimators for average, slope, curvature, change of curvature, etc., of the signal in the time window considered. These coefficients, as well as the approximation error, may be used in a very intuitive way to define segmentation criteria. The properties of SwiftSeg are evaluated by means of some artificial and real benchmark time series. It is compared to three different offline and online techniques to assess its accuracy and runtime. It is shown that SwiftSeg-which is suitable for many data streaming applications-offers high accuracy at very low computational costs.

[1]  Ron Kohavi,et al.  Feature Selection for Knowledge Discovery and Data Mining , 1998 .

[2]  Eamonn J. Keogh,et al.  An online algorithm for segmenting time series , 2001, Proceedings 2001 IEEE International Conference on Data Mining.

[3]  H. Damasio,et al.  IEEE Transactions on Pattern Analysis and Machine Intelligence: Special Issue on Perceptual Organization in Computer Vision , 1998 .

[4]  B. Sick,et al.  Forecasting financial time series with support vector machines based on dynamic kernels , 2008, 2008 IEEE Conference on Soft Computing in Industrial Applications.

[5]  Bernhard Sick,et al.  Processing Short-Term and Long-Term Information With a Combination of Polynomial Approximation Techniques and Time-Delay Neural Networks , 2009, IEEE Transactions on Neural Networks.

[6]  Åke Björck,et al.  Numerical methods for least square problems , 1996 .

[7]  Erich Fuchs.  On Discrete Polynomial Least-Squares Approximation in Moving Time Windows , 1999 .

[8]  Huifeng shen,et al.  Stock tracking : a new multi-dimensional stock forecasting approach , 2005, 2005 7th International Conference on Information Fusion.

[9]  J. Kohlmorgen,et al.  An on-line method for segmentation and identification of non-stationary time series , 2001, Neural Networks for Signal Processing XI: Proceedings of the 2001 IEEE Signal Processing Society Workshop (IEEE Cat. No.01TH8584).

[10]  G. Golub,et al.  Updating and downdating of orthogonal polynomials with data fitting applications , 1991 .

[11]  Steven Lemm,et al.  Fast change point detection in switching dynamics using a hidden Markov model of prediction experts , 1999 .

[12]  Jeffrey M. Hausdorff,et al.  Physionet: Components of a New Research Resource for Complex Physiologic Signals". Circu-lation Vol , 2000 .

[13]  Hongyan Li,et al.  Effective variation management for pseudo periodical streams , 2007, SIGMOD '07.

[14]  Eamonn J. Keogh,et al.  Segmenting Time Series: A Survey and Novel Approach , 2002 .

[15]  Erich Fuchs,et al.  Fast least-squares polynomial approximation in moving time windows , 1997, 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[16]  Jeffrey Scott Vitter.  External memory algorithms , 1998, PODS '98.

[17]  V. B. Uvarov,et al.  Classical Orthogonal Polynomials of a Discrete Variable , 1991 .

[18]  Warwick Tucker,et al.  Foundations of Computational Mathematics a Rigorous Ode Solver and Smale's 14th Problem , 2022 .

[19]  Daniel Lemire,et al.  A Better Alternative to Piecewise Linear Time Series Segmentation , 2006, SDM.

[20]  Patrick Gallinari,et al.  Online Handwritten Shape Recognition Using Segmental Hidden Markov Models , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[21]  Hiroshi Motoda,et al.  Feature Extraction, Construction and Selection: A Data Mining Perspective , 1998 .

[22]  Paul Lukowicz,et al.  Activity Recognition of Assembly Tasks Using Body-Worn Microphones and Accelerometers , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[23]  Hui Ding,et al.  Querying and mining of time series data: experimental comparison of representations and distance measures , 2008, Proc. VLDB Endow..

[24]  Bernhard Sick,et al.  Online Signature Verification with New Time Series Kernels for Support Vector Machines , 2006, ICB.

[25]  Tobias Hanning,et al.  An update algorithm for Fourier coefficients , 2004, 2004 12th European Signal Processing Conference.

[26]  Aiguo Li,et al.  Real-Time Segmenting Time Series Data , 2003, APWeb.

[27]  E. Lorenz.  Deterministic nonperiodic flow , 1963 .

[28]  Robert H. Halstead,et al.  Matrix Computations , 2011, Encyclopedia of Parallel Computing.

[29]  Masamichi Shimosaka,et al.  Online recognition and segmentation for time-series motion with HMM and conceptual relation of actions , 2005, 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[30]  Prabhakar Raghavan,et al.  Computing on data streams , 1999, External Memory Algorithms.

[31]  Paul Lukowicz,et al.  Gesture spotting with body-worn inertial sensors to detect user activities , 2008, Pattern Recognit..

[32]  Pierre-François Marteau,et al.  Time Warp Edit Distance with Stiffness Adjustment for Time Series Matching , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[33]  J. Príncipe,et al.  A fast on-line generalized eigendecomposition algorithm for time series segmentation , 2000, Proceedings of the IEEE 2000 Adaptive Systems for Signal Processing, Communications, and Control Symposium (Cat. No.00EX373).

[34]  Huaiqing Wang,et al.  Novel Online Methods for Time Series Segmentation , 2008, IEEE Transactions on Knowledge and Data Engineering.

[35]  Michel Verhaegen,et al.  ECG Segmentation Using Time-Warping , 1997, IDA.

[36]  Y. Bodyanskiy,et al.  Robust Recursive Fuzzy Clustering-Based Segmentation of Biological Time Series , 2006, 2006 International Symposium on Evolving Fuzzy Systems.

[37]  Daniel Lemire,et al.  Quasi-Monotonic Segmentation of State Variable Behavior for Reactive Control , 2005, AAAI.