Modelling matrix time series via a tensor CP-decomposition

Chang, Jinyuan, Zhang, Henry, Yang, Lin and Yao, Qiwei ORCID: 0000-0003-2065-8486 (2023) Modelling matrix time series via a tensor CP-decomposition. Journal of the Royal Statistical Society. Series B: Statistical Methodology, 85 (1). 127 – 148. ISSN 1369-7412

Text (Modelling Matrix Time Series via a Tensor CP-Decomposition) - Accepted Version Download (571kB)

• Author
• Scopus publication

Abstract

We consider to model matrix time series based on a tensor canonical polyadic (CP)-decomposition. Instead of using an iterative algorithm which is the standard practice for estimating CP-decompositions, we propose a new and one-pass estimation procedure based on a generalized eigenanalysis constructed from the serial dependence structure of the underlying process. To overcome the intricacy of solving a rank-reduced generalized eigenequation, we propose a further refined approach which projects it into a lower-dimensional full-ranked eigenequation. This refined method can significantly improve the finite-sample performance. We show that all the component coefficient vectors in the CP-decomposition can be estimated consistently. The proposed model and the estimation method are also illustrated with both simulated and real data, showing effective dimension-reduction in modelling and forecasting matrix time series.

Item Type:	Article
Official URL:	https://academic.oup.com/jrsssb
Additional Information:	© 2023 (RSS) Royal Statistical Society.
Divisions:	Statistics
Subjects:	H Social Sciences > HA Statistics
Date Deposited:	16 Dec 2022 16:48
Last Modified:	20 Feb 2025 18:04
URI:	http://eprints.lse.ac.uk/id/eprint/117644

Actions (login required)

View Item