我们采用贝叶斯经验贝叶斯方法估计了一个特殊的指数分布族中的刻度参数,证明了刻度参数的贝叶斯经验贝叶斯估计几乎处处收敛到它的贝叶斯估计且获得了它的渐近最优性。最后,提出了一个模拟试验去验证贝叶斯经验贝叶斯估计的渐近最优性。
We employ the Bayes empirical Bayes (EB) procedure to estimate the scale parameter in a special exponential family, and prove that the Bayes EB estimator of the scale parameter converges almost surely (a.s.) to its Bayes estimator and it is asymptotically optimal. Also, a simulation experiment is conducted to demonstrate the asymptotic optimality of the Bayes EB estimator.