doiSerbia

The uniform asymptotic normality of a matrix-T distribution

Li Kai Can (School of Mathematics and Statistics, Hubei Normal University, Huangshi, China)
Zhang Jian Lin (Huangzhou Middle School, Huanggang, China)
Mao Di Shou (School of Mathematics and Statistics, Hubei Normal University, Huangshi, China), likc@hbnu.edu.cn

Using the Kullback-Leibler distance between two density functions about a matrix T distribution and a matrix normal distribution, we obtain a Berry-Esseen boundary for the T distribution. Further, we give the condition under which a matrix T is uniformly asymptotically matrix normal distribution, and point out the convergence rate.

Keywords: matrix T distribution, matrix 􀀀 distribution, Berry-Esseen boundary, uniformly asymptotic normality