本文中,我们利用经验贝叶斯方法研究了指数分布中寿命参数的检验问题。对于假设H0:θ≤θ0←→H1:θ〉θ0,在线性误差损失下,利用两种不同的核估计方法,我们获得了贝叶斯检验风险的同样上界。本文获得的收敛速度优于文献中的早期结果。
We consider the test problem of the life parameter in the exponential distribution using empirical Bayes (EB) approach. For the hypothesis H0 : θ≤ θ0 against H1 : θ〉 θ0, we establish the same upper bound for the regret risk of empirical Bayes test by using two different kernel estimation methods under the linear error loss. The rate of convergence obtained here is better than any other earlier results in the literature.