摘要
在Entropy损失函数下,利用构造多层先验分布的方法求出了指数威布尔分布参数的多层Bayes估计,然后根据经验Bayes估计的思想,利用密度函数的核估计方法,构造了参数的经验Bayes估计并证明了该估计的渐进最优性和可容许性,最后运用随机模拟,将其与平方损失函数下的Bayes估计以及极大似然估计(MLE)进行了比较,结果表明:Entropy损失下的Bayes估计较后两种估计好。
Under the Entropy loss function,the hierarchical Bayes estimation of parameter with exponentiated Weibull distribution has been obtained by means of constructing the hierarchical prior distribution.Then according to empirical Bayes estimation and kernel density estimation,the empirical Bayes estimation of parameter has been constructed and its asymptotic optimality and admissibility have been demonstrated,which has been compared with Bayes estimation under square loss function and MLE.Results show that Bayes estimation under Entropy is better.
出处
《洛阳理工学院学报（自然科学版）》
2014年第4期85-92,共8页
Journal of Luoyang Institute of Science and Technology：Natural Science Edition
基金
甘肃省自然科学基金项目(1208RJZA111)
