Statistical inference of covariance change points in gaussian model

Abstract

In this paper, we study the testing and estimation of multiple covariance change points for a sequence of m-dimensional (m > 1) Gaussian random vectors by using the Schwarz information criterion (SIC). The unbiased SIC is also obtained. The asymptotic null distribution of the test statistic is derived. The result is applied to a simulated bivariate normal vector sequence (m = 2), and changes are successfully detected.

Keywords:

• Change-points
• Information criterion
• SIC
• Asymptotic distribution

Acknowledgement

The authors thank Dr. Y. Gao for simulating the bivariate normal sequence using S-plus. The authors also thank the anonymous referees for their valuable comments.