Bayesian Shrinkage Estimation for Stratified Count Data
In this paper, we consider the problem of simultaneously estimating Poisson parameters under the standardized squared error loss in situations where we can use side information in aggregated data. Bayesian shrinkage estimators are constructed using conjugate gamma and Dirichlet priors. We compare the risk functions of estimators, obtain conditions for domination, and prove minimaxity and admissibility of a proposed estimator. Finally, two extensions are discussed.
READ FULL TEXT