Nonparametric Empirical Bayes Estimation of the Matrix Parameter of the Wishart Distribution

Author: Pensky M.  

Publisher: Academic Press

ISSN: 0047-259X

Source: Journal of Multivariate Analysis, Vol.69, Iss.2, 1999-05, pp. : 242-260

Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.

Previous Menu Next

Abstract

We consider independent pairs (X1, Σ1), (X2, Σ2), …, (Xn, Σn), where each Σi is distributed according to some unknown density function g(Σ) and, given Σi=Σ, Xi has conditional density function q(xΣ) of the Wishart type. In each pair the first component is observable but the second is not. After the (n+1)th observation Xn+1 is obtained, the objective is to estimate Σn+1 corresponding to Xn+1. This estimator is called the empirical Bayes (EB) estimator of Σ. An EB estimator of Σ is constructed without any parametric assumptions on g(Σ). Its posterior mean square risk is examined, and the estimator is demonstrated to be pointwise asymptotically optimal. Copyright 1999 Academic Press.