Asymptotically Optimal Empirical Bayes Estimation for Parameter-functions Of Two-sided Truncation Distribution Families
Ling Zhang · Shuxue de shijian yu renshi · 2000
Consider the two\|sided truncation distribution families written in the form f(x｜θ)=ω(θ 1,θ 2) h(x) I \[θ 1,θ 2\] (x), where θ=(θ 1,θ 2) T(x)=(t\-1(x),t\-2(x))=(min(x\-1,…,x\-m),max(x\-1,…,x\-m)) is a sufficient statistic, its marginal density is denoted by f(t).\;In this paper, by estimating f(t), we construct the empirical Bayes estimation (EBE) for parameter\|function Q(θ), and prove the EBE is an asymptotically optimal that of Q(θ).