逐步增加的Ⅱ型截尾下Lomax分布形状参数的估计  

Bayesian Estimation of the Shape Parameter for Lomax Distribution Based on Progressively Type-Ⅱ Censored Tests

作　　者：龙兵[1] 王芳[1] 习长新[1] 

机构地区：[1]荆楚理工学院数理学院,湖北荆门448000

出　　处：《统计与决策》2017年第8期15-18,共4页Statistics & Decision

基　　金：湖北省教育厅科研资助项目(B2016264);荆楚理工学院院级科研资助项目(ZR201504)

摘　　要：文章在逐步增加的Ⅱ型截尾下,给出了Lomax分布形状参数θ的极大似然估计;由"平均剩余寿命"的概念得到了形状参数的逆矩估计,在平方损失函数和对称熵损失函数下,针对不同的先验分布给出了参数θ的Bayes估计;最后通过随机模拟对几个估计进行了比较,说明了在相同的损失函数下,取共轭先验分布较无信息先验分布的精度要高。This paper offers maximum likelihood estimation of the shape parameter for Lomax distribution under the progres- sively type-II censored samples. And then, inverse moment estimation of the shape parameter is derived from concept of ＂average remaining life＂, and aiming at different prior distributions, the Bayesian estimation of the parameter [ 0 ] is given under the condition of squared loss function and symmetric entropy loss function. Finally, a few estimations are compared through random simulations and the result indicates that the conjugate prior distribution has higher precession with the same loss function.

关 键 词：Lomax分布 逐步增加的Ⅱ型截尾 逆矩估计 先验分布 BAYES估计 

分 类 号：O212.8[理学—概率论与数理统计]