Open Access
2003 NON-CENTRAL MATRIX-VARIATE DIRICHLET DISTRIBUTION
Luz Estela S´anchez, Daya K. Nagar
Taiwanese J. Math. 7(3): 477-491 (2003). DOI: 10.11650/twjm/1500558399

Abstract

Let $X_{i}\sim W_{p}(n_{i},\Sigma,\Theta)$ where $\Theta= diag(\theta^{2}_{i},0,\ldots,0)$, $i=1,\ldots,r+1$. In this article the authors have derived the joint distribution of $U_i=C^{-1}X_{i}C'{}^{-1}$, $i=1,\ldots,r$ where $\sum_{i=1}^{r+1}X_{i}=CC'$ and $C$ is a lower triangular matrix. The joint distribution of $U_{1},\ldots,U_{r}$ is a non-central matrix-variate Dirichlet distribution. Several properties of this distribution such as marginal and conditional distributions, distribution of partial sums, moments and asymptotic results have also been studied.

Citation

Download Citation

Luz Estela S´anchez. Daya K. Nagar. "NON-CENTRAL MATRIX-VARIATE DIRICHLET DISTRIBUTION." Taiwanese J. Math. 7 (3) 477 - 491, 2003. https://doi.org/10.11650/twjm/1500558399

Information

Published: 2003
First available in Project Euclid: 20 July 2017

MathSciNet: MR1998769
Digital Object Identifier: 10.11650/twjm/1500558399

Subjects:
Primary: 62E15
Secondary: 62H99

Keywords: asymptotic , confluent hypergeometric function , Dirichlet distribution , Kamp\'e de F\'eriet's function , matrix variate , matrix-variate beta distribution , non-central , transformation

Rights: Copyright © 2003 The Mathematical Society of the Republic of China

Vol.7 • No. 3 • 2003
Back to Top