一种对称损失下两参数广义Pareto分布形状参数的Bayes分析  

Bayes estimation of the shape parameters of the generalized Pareto distribution with two parameters under a symmetric loss function

作　　者：吴月丹 徐宝[1] WU Yuedan;XU Bao(College of Mathematics and Computer Science,Jilin Normal University,Siping Jilin 136000,China)

机构地区：[1]吉林师范大学数学与计算机学院,吉林四平136000

出　　处：《南昌大学学报（理科版）》2023年第1期21-27,共7页Journal of Nanchang University(Natural Science)

基　　金：国家自然科学基金资助项目(11571138);吉林省科技发展计划项目(YDZJ202201ZYTS622)。

摘　　要：在加权对称损失函数下,使用Bayes估计方法,讨论了两参数广义Pareto分布形状参数的Bayes估计及其性质,得到了形状参数的Bayes估计的一般形式与在两种不同先验下的精确形式,并由此得到形状参数的最小最大估计的一种形式,将不同的估计联系在一起。并通过模拟研究了所得估计的精度,结果表明所得估计的精度较高。For the weighted symmetric loss function, the Bayes estimation method was used to study the Bayes estimation of the shape parameter of two-parameter generalized Pareto distribution and its properties.The general form of Bayes estimation of the shape parameters and the exact form under two different prior distributions were obtained, and a form of the minimax estimation of the shape parameter was obtained.So the connection among different estimates were established.The accuracy of the estimate was studied by simulation, and the result showed that the accuracy of the estimate is higher.

关 键 词：广义PARETO分布 BAYES估计 损失函数 可容许性 最小最大估计 

分 类 号：O212.8[理学—概率论与数理统计]