连续型单参数指数族参数的经验Bayes估计问题:NA样本情形

The Empirical Bayes Estimation Problem for Continuous One-parameter Exponential Family:In the case of NA Samples

对连续型单参数指数族在平方损失下导出了参数的Bayes估计,利用同分布负相协(NA)样本构造了经验Bayes(EB)估计量,并在适当条件下获得了EB估计的收敛速度.文末给出一个满足定理条件的例子.

In this paper, the Bayes estimation problem for the continuous one-parameter exponential family has heen derived and the empirical Bayes （EB） estimator is constructed hy using the identicslly disttihuted and negatively associated （NA） samples. It is shown that the proposed EB estimator is asymptotically optimal with convergence rate near O（n^-1/2） under suitable conditions. Finally, an example concerning main resuh is given.