Dynamic Time Warping (DTW) is considered as a robust measure to compare numerical time series when some time elasticity is required. However, speed is a known major drawback of DTW due to its quadratic complexity. Previous work has mainly considered designing speed optimization based on early-abandoning strategies applied to nearest-neighbor classification, although some of these optimizations are restricted to uni-dimensional time series. In this paper, we introduce Coarse-DTW, a reinterpretation of DTW for sparse time series, which exploits adaptive downsampling to achieve speed enhancement, even when faced with multidimensional time series. We show that Coarse-DTW achieves nontrivial speedups in nearest-neighbor classification and even admits a positive-definite kernelization suitable for SVM classification, hence offering a good tradeoff between speed and accuracy.
CITATION STYLE
Dupont, M., & Marteau, P. F. (2016). Coarse-DTW for sparse time series alignment. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9785 LNCS, pp. 157–172). Springer Verlag. https://doi.org/10.1007/978-3-319-44412-3_11
Mendeley helps you to discover research relevant for your work.