Similarity Measure for Multivariate Time Series Based on Dynamic Time Warping

Similarity measure for multivariate time series is a hot topic in the area of data mining. However, existing algorithms of similarity measure cannot resolve the contradiction between matching accuracy and computational complexity. We propose a novel similarity measure for multivariate time series. First, important points are extracted from multivariate time series. Then, a similarity measure based on dynamic time warping is proposed. Finally, the performance of our proposed method and other popular approaches is compared. The experimental results show that the proposed method can effectively measure the similarity of multivariate time series at relatively lower computational cost.