The paper “Modeling Matrix Time Series via a Tensor CP-Composition”, co-authored by Prof.Jinyuan Chang, Jing He, and doctoral student Lin Yang from the Joint Laboratory of Data Science and Business Intelligence at Southwestern University of Finance and Economics, along with Prof. Qiwei Yao from the London School of Economics and Political Science, has been officially accepted by the Journal of the Royal Statistical Society: Series B, one of the top international journals in the field of statistics.
Abstract
This paper uses tensor CP-decomposition to model matrix time series and proposes using the autocorrelation information of observation sequences to construct generalized eigenvalue equations for estimating CP decomposition. This method is computationally simple and does not require iterations, overcoming the drawback of traditional CP decomposition estimation methods that require multiple iterations to converge. In order to solve the computational uncertainty of solving the non rank generalized eigenvalue equation, this paper also proposes an improved method that transforms it into solving a low dimensional full rank generalized eigenvalue equation, effectively improving the estimation performance under finite samples. Numerical simulation and case analysis show that the newly proposed method has excellent performance in modeling and predicting matrix time series.
Author Introduction
Jinyuan Chang is the Executive Director of the Joint Laboratory of Data Science and Business Intelligence at SWUFE. He is a Guanghua Distinguished Professor, Doctoral Supervisor, recipient of the National Science Fund for Distinguished Young Scholars of China, a Distinguished Expert of Sichuan Province, and a member of the Sichuan Province Statistical Expert Advisory Committee. His research primarily focuses on ultra-high dimensional data analysis and high-frequency financial data analysis.
Jing He is an Associate Professor at the Joint Laboratory of Data Science and Business Intelligence at SWUFE. Her research areas include high-dimensional data analysis and spatiotemporal data analysis.
Lin Yang is a Ph.D. student in statistics at the School of Statistics, SWUFE, under the supervision of Professor Jinyuan Chang. His research focuses on ultra-high dimensional time series analysis and functional time series analysis.
Qiwei Yao is a Chair Professor at the London School of Economics and Political Science. He is a Fellow of the Institute of Mathematical Statistics, an Elected Member of the International Statistical Institute, and a Fellow of the American Statistical Association. His research covers a wide range of topics including time series analysis, high-dimensional time series modeling and prediction, dimensionality reduction, factor modeling, dynamic network modeling, spatiotemporal modeling, financial econometrics, and nonparametric regression.
