Accelerating Dynamic Time Warping Clustering with a Novel Admissible Pruning Strategy

Abstract

Clustering time series is a useful operation in its own right, and an important subroutine in many higher-level data mining analyses, including data editing for classifiers, summarization, and outlier detection. While it has been noted that the general superiority of Dynamic Time Warping (DTW) over Euclidean Distance for similarity search diminishes as we consider ever larger datasets, as we shall show, the same is not true for clustering. Thus, clustering time series under DTW remains a computationally challenging task. In this work, we address this lethargy in two ways. We propose a novel pruning strategy that exploits both upper and lower bounds to prune off a large fraction of the expensive distance calculations. This pruning strategy is admissible; giving us provably identical results to the brute force algorithm, but is at least an order of magnitude faster. For datasets where even this level of speedup is inadequate, we show that we can use a simple heuristic to order the unavoidable calculations in a most-useful-first ordering, thus casting the clustering as an anytime algorithm. We demonstrate the utility of our ideas with both single and multidimensional case studies in the domains of astronomy, speech physiology, medicine and entomology.

Interactive time series exploration powered by the marriage of similarity distances

“…For example, a benchmark dataset like StarLightCurves from the UCR Time Series Data Mining Archive 2 with 9236 time series, each of length 1024, is composed of 4.83e 9 subsequences. Real-world datasets tend to be orders of magnitude larger than this 3 . Performing similarity comparisons for all these subsequences is clearly impractical.…”

“…[21] does data editing to make its classifier faster and more accurate, while we have to keep the original data to support accurate similarity searches and return actual sequence results. [3] introduces clusters constructed using DTW and a modified Density Peaks algorithm to improve the clustering performance. Using upper and lower bounds of DTW, TADPole can prune unnecessary computations.…”

Battery Grouping with Time Series Clustering Based on Affinity Propagation

“…Energies 2016, 9, 561 6 of 11 DTW utilizes dynamic programming to find the optimal mapping of points in two sequences and compute the distance between them. Detailed steps for computing the DTW distance can be found in several papers [20,21], and efficient algorithms can be found in [17].…”

“…The battery charge/discharge sequences are first denoised. The dynamic time warping (DTW) distances [17] between different battery charge/discharge sequences are then calculated, followed by a normalization step to form a similarity matrix, which is utilized by the affinity propagation (AP) algorithm [18] for clustering.…”

Video Shots‘ Matching via Various Length of Multidimensional Time Sequences

“…There are lots of different methods nowadays for solving the task. A special attention in clustering data arrays is given to issues related to time series processing, which usually imply (except for actual clustering) segmentation, detection of property changes (fault detection) and abnormal outliers [13]. Thereby, clustering should be performed not only in space, but also in time.…”