Amercing: An Intuitive, Elegant and Effective Constraint for Dynamic Time Warping

Dynamic Time Warping (DTW), and its constrained (CDTW) and weighted (WDTW) variants, are time series distances with a wide range of applications. They minimize the cost of non-linear alignments between series. CDTW and WDTW have been introduced because DTW is too permissive in its alignments. However, CDTW uses a crude step function, allowing unconstrained flexibility within the window, and none beyond it. WDTW’s multiplicative weight is relative to the distances between aligned points along a warped path, rather than being a direct function of the amount of warping that is introduced. In this paper, we introduce Amerced Dynamic Time Warping (ADTW), a new, intuitive, DTW variant that penalizes the act of warping by a fixed additive cost. Like CDTW and WDTW, ADTW constrains the amount of warping. However, it avoids both abrupt discontinuities in the amount of warping allowed and the limitations of a multiplicative penalty. We formally introduce ADTW, prove some of its properties, and discuss its parameterization. We show on a simple example how it can be parameterized to achieve an intuitive outcome, and demonstrate its usefulness on a standard time series classification benchmark. We provide a demonstration application in C++ [1].

[1]  Jason Lines,et al.  Time series classification with ensembles of elastic distance measures , 2015, Data Mining and Knowledge Discovery.

[2]  Olufemi A. Omitaomu,et al.  Weighted dynamic time warping for time series classification , 2011, Pattern Recognit..

[3]  Eamonn J. Keogh,et al.  Time series motifs discovery under DTW allows more robust discovery of conserved structure , 2021, Data Mining and Knowledge Discovery.

[4]  Gunasekaran Manogaran,et al.  Wearable sensor devices for early detection of Alzheimer disease using dynamic time warping algorithm , 2018, Cluster Computing.

[5]  P. Alam.  ‘L’ , 2021, Composites Engineering: An A–Z Guide.

[6]  Pong C. Yuen,et al.  Invariant subspace learning for time series data based on dynamic time warping distance , 2020, Pattern Recognit..

[7]  Gurdit Singh,et al.  Smart patrolling: An efficient road surface monitoring using smartphone sensors and crowdsourcing , 2017, Pervasive Mob. Comput..

[8]  Geoffrey I. Webb,et al.  TS-CHIEF: a scalable and accurate forest algorithm for time series classification , 2019, Data Mining and Knowledge Discovery.

[9]  Pierre Gançarski,et al.  A global averaging method for dynamic time warping, with applications to clustering , 2011, Pattern Recognit..

[10]  Hong Cheng,et al.  An image-to-class dynamic time warping approach for both 3D static and trajectory hand gesture recognition , 2016, Pattern Recognit..

[11]  Eamonn J. Keogh,et al.  Searching and Mining Trillions of Time Series Subsequences under Dynamic Time Warping , 2012, KDD.

[12]  Zahraa Yasseen,et al.  Shape matching by part alignment using extended chordal axis transform , 2016, Pattern Recognit..

[13]  Geoffrey I. Webb,et al.  Proximity Forest: an effective and scalable distance-based classifier for time series , 2018, Data Mining and Knowledge Discovery.

[14]  Geoffrey I. Webb,et al.  Early abandoning and pruning for elastic distances including dynamic time warping , 2021, Data Min. Knowl. Discov..

[15]  Geoffrey I. Webb,et al.  Time series extrinsic regression , 2020, Data mining and knowledge discovery.

[16]  Yanfei Kang,et al.  Improving the Accuracy of Global Forecasting Models using Time Series Data Augmentation , 2020, Pattern Recognit..

[17]  Geoffrey I. Webb,et al.  FastEE: Fast Ensembles of Elastic Distances for time series classification , 2019, Data Mining and Knowledge Discovery.

[18]  Eamonn J. Keogh,et al.  The UCR time series archive , 2018, IEEE/CAA Journal of Automatica Sinica.

[19]  F. Itakura,et al.  Minimum prediction residual principle applied to speech recognition , 1975 .

[20]  Janez Demsar,et al.  Statistical Comparisons of Classifiers over Multiple Data Sets , 2006, J. Mach. Learn. Res..

[21]  Edwin C. Kan,et al.  A real-time spike classification method based on dynamic time warping for extracellular enteric neural recording with large waveform variability , 2016, Journal of Neuroscience Methods.

[22]  Sangarapillai Lambotharan,et al.  Anomaly Detection Using Dynamic Time Warping , 2019, 2019 IEEE International Conference on Computational Science and Engineering (CSE) and IEEE International Conference on Embedded and Ubiquitous Computing (EUC).

[23]  S. Chiba,et al.  Dynamic programming algorithm optimization for spoken word recognition , 1978 .

[24]  Manabu Okawa,et al.  Online signature verification using single-template matching with time-series averaging and gradient boosting , 2020, Pattern Recognit..