A Similarity Measurement for Multivariate Time Series Based on Variable Clustering

Abstract

The existing multivariate time series similarity measurement methods cannot achieve high accuracy and high efficiency at the same time. To solve this problem, a similarity measurement method of multivariate time series based on variable clustering is proposed. OPTICS clustering is used to cluster highly correlated variables, and PAA is used to obtain the feature sequences by reducing the scale of time dimension, the feature sequences are imported to dynamic time warping to measure the similarity of multivariate time series. Taking public data sets as experiment data, the results indicate that the proposed method improves the efficiency of similarity measurement and the precision of similar matching to a certain extent.