Convergence rates for empirical Bayes estimation of binomial parameter (Q1117616)

This paper deals with the problem of estimating the binomial parameter through the empirical Bayes approach. Two empirical Bayes estimators are considered. We prove that the associated Bayes risks of the concerned estimators converge to the minimum Bayes risk at least with rate \(O(n^{-1}\log n)\), where n is the number of accumulated post experience at hand. An asymptotically lower bound for the convergence rate is also found. This lower bound converges to zero with rate \(O(n^{-1})\).