双边定数截尾下Pareto分布的Bayes估计

The Bayesian Estimation for the Pareto Distribution under Type-Ⅱ Doubly Censored Sample

基于双边定数截尾样本,在LINEX损失函数和复合LINEX损失函数下,考虑无信息先验分布与共轭先验分布,讨论Pareto分布的形状参数的Bayes估计,并对各条件下参数的估计进行数值模拟比较分析。结果表明,在不同损失函数下,取共轭先验分布时参数的估计优于无信息先验分布下的估计,且在LINEX损失函数下,先验分布取Γ分布时,参数估计的均方误差较小,估计的效果更好。

Based on Type-Ⅱ Doubly Censored Sample,under the LINEX loss function and composite LINEX loss function,considering the uninformative priori distribution and conjugate priori distribution,this paper discusses the Bayesian estimation of the shape parameters of Pareto distribution,and makes numerical simulation and comparative analysis on the estimation of parameters under various conditions.The results show that under different loss functions,the estimation of parameters under conjugate priori distribution is better than that under uninformative a priori distribution,and under the LINEX loss function,when the priori distribution takes theΓdistribution,the mean square error of parameter estimation is smaller and the estimation effect is better.