熵损失下负二项分布可靠度的E-Bayes估计  

E-Bayesian estimation of the negative binomial distribution reliability under entropy loss function

作　　者：苏清华[1,2] 刘次华[2] 

机构地区：[1]孝感学院数学系,湖北孝感432000 [2]华中科技大学数学与统计学院,湖北武汉430074

出　　处：《湖北师范学院学报（自然科学版）》2009年第3期14-17,共4页Journal of Hubei Normal University(Natural Science)

基　　金：湖北省教育厅优秀中青年人才项目(Q200726003)资助;孝感学院科研项目(Z2009002)资助

摘　　要：对负二项分布,给出可靠度在熵损失函数下的Bayes估计,在此基础上给出当先验分布为幂分布时,可靠度的E-Bayes估计和多层Bayes估计,并设计了相应的数值实验,实验结果表明负二项分布可靠度的E-Bayes估计较多层Bayes估计稳健性更好,计算更简单,便于实际应用.In this paper, for negative Binomial distribution, the Bayesian estimation of reliability under entropy loss function is given, and on the base, the E - Bayesian estimation and hierarchical Bayesian estimation of 0 are given when the prior distribution of θ is power distribution. Finally, the numerical results shows that the E - Bayesian estimation of θ is more robust and easier to calculate than its hierarchical Bayesian estimation, that is, the E - Bayesian estimation of θ is easier to operate.

关 键 词：负二项分布 熵损失函数 幂分布 E—Bayes估计 多层BAYES估计 

分 类 号：O213.2[理学—概率论与数理统计]