Gao Jiti
Journal of Systems Science and Complexity. 1989, 2(4): 317-324.
In this paper, we obtain empirical Bayes(EB)procedures for selecting the best among k different exponential populations(the form of the conditional probability density e.g. each population is f_4(x_i/b_i)=b_i exp(-b_ix_i),x_i,b_i∈(0,∞),i=1,2,…,k).These rules are based on the EB estimators of b_i. We show that, under the squared error loss, the Bayes risk of the EB estimators converges to the related minimum Bayes risks with rates of conver-gence at Jeast of order O(n~(-q)). Further, for the selection problem, the rates of convergence of the proposed selection rules are shown to be at least of order O(n~((-q)/2) where q can be arbitrarily close to 1/5 or 1 under suitable conditions.