A Total Variation Based Method for Multivariate Time Series Segmentation

作     者：Min Li Yumei Huang Youwei Wen 

作者机构：School of Mathematics and StatisticsCenter for Data ScienceLanzhou UniversityLanzhouGansu 730000China School of Mathematics and StatisticsHunan Normal UniversityChangshaHunan 410081China and Key Laboratory of Computing and Stochastic Mathematics(LCSM)Ministry of Education of China 

出 版 物：《Advances in Applied Mathematics and Mechanics》 (应用数学与力学进展（英文）)

年 卷 期：2023年第15卷第2期

页      面：300-321页

核心收录：

学科分类：07[理学] 0701[理学-数学] 0801[工学-力学（可授工学、理学学位）] 070101[理学-基础数学] 

基　　金：This work is supported by the National Natural Science Foundation of China Nos.11971215,11871210,and 11971214 the Key Laboratory of Applied Mathematics and Complex Systems of Lanzhou University 

主　　题：Multivariate time series segmentation total variation dynamic programming 

摘      要：Multivariate time series segmentation is an important problem in data mining and it has arisen in more and more practical applications in recent *** task of time series segmentation is to partition a time series into segments by detecting the abrupt changes or anomalies in the time *** time series segmentation can provide meaningful information for further data analysis,prediction and policy decision.A time series can be considered as a piecewise continuous function,it is natural to take its total variation norm as a prior information of this time *** this paper,by minimizing the negative log-likelihood function of a time series,we propose a total variation based model for multivariate time series *** iterative process is applied to solve the proposed model and a search combined the dynamic programming method is designed to determine the *** experimental results show that the proposed method is efficient for multivariate time series segmentation and it is competitive to the existing methods for multivariate time series segmentation.