多变量时间序列聚类综述

A review of multivariate time series clustering

多变量时间序列(Multivariate Time Series,MTS)具有多变量性和高冗余性,使用聚类分析从海量、高维的MTS数据中挖掘有趣模式具有重要意义。本文从基于实例、基于特征和基于模型的角度,对近年来MTS聚类方法的研究进行归类,为研究者了解最新的MTS聚类方法研究动态和发展趋势提供参考。

Multivariate time series is multivariable and highly redundant.Cluster analysis is of great significance in mining interesting patterns from massive and high-dimensional MTS data.In this paper,the research on MTS clustering methods in recent years is classified from the perspectives of case-based,feature-based and model-based,so as to provide reference for researchers to understand the latest research trends and development trends of MTS clustering methods.