Minimax Estimation of a Multivariate Normal Mean with Unknown Covariance Matrix.
Abstract
Let X be a p-variate (p>or=3) vector, normally distributed with unknown mean theta and unknown covariance matrix sigma. Let W:pXp be distributed independently of X, and let W have a Wishart distribution with n degrees of freedom and parameter sigma. It is desired to estimate theta under the quadratic loss (delta-theta)'Q(delta-theta), where Q is a known positive definite matrix. Under the condition that a lower bound for the smallest characteristic root of Q sigma is known, a family of minimax estimators is developed.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1976
- Accession Number
- ADA037531
Entities
People
- Leon Jay Gleser
Organizations
- Purdue University