双边定时截尾下Pareto分布的参数的极大似然估计的EM算法  

The EM Algorithm of the Maximum Likelihood Estimatiom of Pareto Distribution under Type-I Doubly Censored Sample

作　　者：田霆[1] 刘次华[2] TIAN Ting;LIU Cihua(School of Mathematical Sciences,Hua Qiao University,Quanzhou 362021,China;School of Mathematics and Statistics,HUST,Wuhan 430074,China)

机构地区：[1]国立华侨大学数学科学学院,福建泉州362021 [2]华中科技大学数学与统计学院,湖北武汉430074

出　　处：《电子产品可靠性与环境试验》2024年第3期52-54,共3页Electronic Product Reliability and Environmental Testing

摘　　要：给出了当寿命分布为Pareto分布时,双边定时截尾寿命试验下形状参数的极大似然估计。由于似然方程形式较复杂,无法得到参数的显式表达式。但可证明此极大似然估计是唯一存在的,并利用EM算法求出了此参数的一种估计。The maximum likelihood estimation of shape parameters under the bilateral type-I censoring life test is given when the life distribution is Pareto.Due to the complexity of the likelihood equation,the explicit expression of parameter estimation can not be obtained.However,it can be shown that this maximum likelihood estimate is the only one,and EM algorithm is used to find an estimate of this parameter.

关 键 词：PARETO分布 双边定时截尾 极大似然估计 EM算法 

分 类 号：O213.2[理学—概率论与数理统计]