EM算法的一个应用  

An application of EM algorithm

作　　者：温艳清[1] 

机构地区：[1]山西大同大学数学与计算机科学学院,山西大同037009

出　　处：《山东理工大学学报（自然科学版）》2012年第6期66-68,共3页Journal of Shandong University of Technology:Natural Science Edition

摘　　要：介绍了EM算法的概念、步骤和性质.在不完全数据条件下,对Weibull分布使用EM算法进行极大似然估计,并且在得到相应的迭代公式后,进行了随机模拟.模拟结果证明了利用该算法处理Weibull分布参数的极大似然估计的优良性.This paper briefly introduces the concept, steps and nature of EM algorithm. The maximum likelihood estimation of the Weibull distribution was produced using the EM algorithm under the condition of incomplete data, the corresponding iterative formula was obtained, and the stochastic simulation was carried out. The simulation results illustrate that the algorithm is good at dealing with maximum likelihood estimation of Weibull distribution parameters.

关 键 词：不完全数据 极大似然估计 EM算法 

分 类 号：O213[理学—概率论与数理统计] O211.62[理学—数学]