摘要
在逐步Ⅰ型区间删失数据下,试图求Lomax分布中未知形状参数的极大似然估计,事实上并不能得到参数的显式表达式.根据形状参数的后验密度函数,用EM算法得到了形状参数的迭代公式,两个引理说明EM算法具有良好的收敛性.通过随机模拟的例子说明了EM算法的可行性,并且最终估计值与初值的选取无关.
An attempt is made to obtain the maximum likelihood estimation of the unknown shape parameter in Lomax distribution under progressive type-I interval censored data,in fact,the explicit expression of the parameter can not be obtained.The iterative formula of the shape parameter is obtained by using the EM algorithm based on the posterior density function of the shape parameter,the two lemmas show that the EM algorithm has good convergence.The feasibility of the EM algorithm is illustrated by the example of stochastic simulation,and the final estimation is independent of the initial value.
作者
龙沁怡
唐俊
LONG Qinyi;TANG Jun(School of Science,Inner Mongolia University of Science and Technology,Baotou 014010,China)
出处
《周口师范学院学报》
CAS
2018年第2期1-4,共4页
Journal of Zhoukou Normal University
基金
内蒙古自治区教育厅科研项目(No.NJZY17168)
关键词
Lomax分布
逐步I型区间删失
极大似然估计
EM算法
Lomax distribution
progressive type-I interval censoring
maximum likelihood estimation
EM algorithm
