EA_DTW: Early Abandon to Accelerate Exactly Warping Matching of Time Series

摘要

Dynamic Time Warping (DTW) is one of the im- portant distance measures for time series, however,the exact calculation of DTW has become a bot-tleneck. We propose an approach, named Early Abandon DTW (EA DTW) to accelerate the cal-culation. We demonstrate the idea of early aban-don by theoretical analysis, and show the utili-ties of EA DTW by thorough experiments both on synthetic and real datasets. The results show,EA DTW outperforms the dynamic DTW calcula-tion in the light of process time, and is much bet-ter when the threshold is below the real DTW dis-tance.