Empirical and hierarchical Bayes competitors of preliminary test estimators in two sample problems
Empirical and hierarchical Bayes competitors of preliminary test estimators in two sample problems
复制标题
两个样本问题中初步测试估计器的经验和分层贝叶斯竞争者
DOI：
10.1016/0047-259x(88)90126-1
复制
发表时间：
1988
影响因子：
1.6
通讯作者：
B. K. Sinha
中科院分区：
文献类型：
--
作者：
M. Ghosh;B. K. Sinha
考虑了基于np （μ1,σ2V1）和np （μ2,σ2V2）的独立样本怀疑μ1≈μ2时μ1的估计问题。我们假设v1， v2已知，但σ2未知。首先，推导了EB估计量，研究了其贝叶斯性质和频域性质。其次，提出了一个改进的EB估计量，并证明其优于一个初步测试估计量。最后，提出了一种层次贝叶斯方法作为EB估计的替代方法。
We consider the problem of estimation ofμ1when it is suspected thatμ1≈μ2based on independent samples fromNp(μ1,σ2V1) andNp(μ2,σ2V2). We assumeV1,V2known butσ2unknown. First, the EB estimator is derived and its Bayesian and frequentist properties are studied. Second, a modified EB estimator is proposed and shown to dominate a preliminary test estimator. Finally, a hierarchical Bayes approach is proposed as an alternative to EB estimators.