缺失数据场合下Frechet分布参数的逆矩估计  

Inverse Moment Estimates for the Parameters of Frechet Distribution with Missing Data

作　　者：何亮 徐晓岭 吴生荣 HE Liang;XU Xiaoling;WU Shengrong(School of Statistics and Information,Shanghai University of International Business and Economics,Shanghai 201620,China;Yixing Entry-Exit Inspection&Quarantine Bureau,Yixing 214206,China)

机构地区：[1]上海对外经贸大学统计与信息学院,上海201620 [2]宜兴出入境检验检疫局,江苏宜兴214206

出　　处：《兵器装备工程学报》2018年第10期197-203,共7页Journal of Ordnance Equipment Engineering

基　　金：上海对外经贸大学2017年研究生科研创新项目

摘　　要：给出了Frechet分布在缺失数据场合下两种逆矩估计。通过Monte-Carlo模拟试验及算例比较近似极大似然估计与逆矩估计的结果,表明了在样本数据缺失下逆矩估计效果优于近似极大似然估计,并对两种逆矩估计的结果进行了比较。Two methods of estimating the parameters of the two-parameter Frechet distribution with missing data was given.A comparison between the results of approximated maximum likelihood estimation and inverse moment estimate was performed through monte-carlo simulations and some examples.The results show that the inverse moment estimate was superior to approximatedd maximum likelihood estimation under the circumstances of missing sample data.And it determined which method is better between the two inverse moment estimates.

关 键 词：Frechet分布 缺失数据 失效率函数 逆矩估计 均方误差 

分 类 号：O213.2[理学—概率论与数理统计]