Abstract
Dynamic time warping (DTW) has been applied to a wide range of machine learning problems involving the comparison of time series. An important feature of such time series is that they can sometimes be sparse in the sense that the data takes zero value at many epochs. This corresponds for example to quiet periods in speech or to a lack of physical or dietary activity. However, employing conventional DTW for such sparse time series runs a full search ignoring the zero data.
In this paper we focus on the development and analysis of a fast dynamic time warping algorithm that is exactly equivalent to DTW for the unconstrained case where there are no global constraints on the permissible warping path. We call this sparse dynamic time warping (SDTW). A careful formulation and analysis are performed to determine exactly how SDTW should treat the zero data. It is shown that SDTW reduces the computational complexity relative to DTW by about twice the sparsity ratio, which is defined as the arithmetic mean of the fraction of non-zero’s in the two time series. Numerical experiments confirm the speed advantage of SDTW relative to DTW for sparse time series with sparsity ratio up to 0.4. This study provides a benchmark and also background to potentially understand how to exploit such sparsity in the more complex case of a global constraint, or when the underlying time series are approximated to reduce complexity.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Assent, I., Wichterich, M., Krieger, R., Kremer, H., Seidl, T.: Anticipatory DTW for efficient similarity search in time series databases. Proc. VLDB Endowment 2(1), 826–837 (2009)
Berndt, D.J., Clifford, J.: Using dynamic time warping to find patterns in time series. In: Knowledge Discovery in Databases Workshop, Seattle, WA, vol. 10, pp. 359–370 (1994)
Esling, P., Agon, C.: Time-series data mining. ACM Comput. Surv. (CSUR) 45(1), 12 (2012)
Tak-chung, F.: A review on time series data mining. Eng. Appl. Artif. Intell. 24(1), 164–181 (2011)
Grabocka, J., Nanopoulos, A., Schmidt-Thieme, L.: Classification of sparse time series via supervised matrix factorization. In: AAAI (2012)
Itakura, F.: Minimum prediction residual principle applied to speech recognition. IEEE Trans. Acoust. Speech Signal Process. 23(1), 67–72 (1975)
Eamonn Keogh and Chotirat Ann Ratanamahatana: Exact indexing of dynamic time warping. Knowl. Inf. Syst. 7(3), 358–386 (2005)
Keogh, E., Xi, X., Wei, L., Ratanamahatana, C.A.: The UCR time series classification/clustering homepage (2006). http://www.cs.ucr.edu/~eamonn/time_series_data
Keogh, E.J., Pazzani, M.J. Scaling up dynamic time warping for datamining applications. In: Proceedings of the Sixth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 285–289. ACM (2000)
Müller, M.: Information Retrieval for Music and Motion, vol. 2. Springer, Heidelberg (2007)
National Centers for Health Statistics, Centers for Disease Control, Prevention. National health, nutrition examination survey data 2011–2012. Centers for Disease Control and Prevention, Hyattsville, MD, US Department of Health and Human Services (2014)
Sakoe, H., Chiba, S.: Dynamic programming algorithm optimization for spoken word recognition. IEEE Trans. Acoust. Speech Signal Process. 26(1), 43–49 (1978)
Salvador, S., Chan, P.: Toward accurate dynamic time warping in linear time and space. Intell. Data Anal. 11(5), 561–580 (2007)
Zhao, Y., Hwang, Y., Gelfand, S., Delp, E., Eicher-Miller, H.: Temporal dietary patterns derived from spectral clustering and factor analysis are associated with diet quality using nhanes 1999–2004. FASEB J. 29(1 Supplement), 587–19 (2015)
Zhou, M., Wong, M.H.: Boundary-based lower-bound functions for dynamic time warping and their indexing. Inf. Sci. 181(19), 4175–4196 (2011)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Hwang, Y., Gelfand, S.B. (2017). Sparse Dynamic Time Warping. In: Perner, P. (eds) Machine Learning and Data Mining in Pattern Recognition. MLDM 2017. Lecture Notes in Computer Science(), vol 10358. Springer, Cham. https://doi.org/10.1007/978-3-319-62416-7_12
Download citation
DOI: https://doi.org/10.1007/978-3-319-62416-7_12
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-62415-0
Online ISBN: 978-3-319-62416-7
eBook Packages: Computer ScienceComputer Science (R0)