Volume no :2, Issue no: 2, December (2009)

Journal of Statistics: Advances in Theory and Applications

REDUCED GRADIENT METHOD FOR MINIMAX ESTIMATION OF A BOUNDED POISSON MEAN

Author's: A. El Mouatasim and A. Al-Hossain
Pages: [185] - [199]
Received Date: September 30, 2009
Submitted by:

Abstract

This paper is concerned with minimax estimation of a bounded mean Poisson given the prior information that the mean lies in a bounded domain, and using the information normalized squared error loss. It would only be necessary to find the correct corresponding Bayes estimator, and we present the algorithm of reduced gradient via stochastic perturbation for solving this statistical problem.

Keywords

Bayes risk, loss estimation, minimax risk, global optimization.