具有部分缺失数据混合指数分布参数的估计  

Partially Missing Data Estimation of Parameters of Mixed Exponential Distribution

作　　者：李阳[1] 杨艳秋[1] 王秋爽 徐圣楠 赵志文[1] LI Yang;YANG Yan-qiu;WANG Qiu-shuang;XU Sheng-nan;ZHAO Zhi-wen(College of Mathematics,Jilin Normal University,Siping 136000,China)

机构地区：[1]吉林师范大学数学学院,吉林四平136000

出　　处：《白城师范学院学报》2018年第8期10-15,共6页Journal of Baicheng Normal University

基　　金：国家自然科学基金项目(11571138)

摘　　要：研究具有部分缺失数据混合指数分布总体参数的估计与检验问题.一方面,给出了未知参数的矩估计,同时证明了估计量的相合性以及渐近正态性.另一方面,我们讨论了两个混合指数分布总体参数相等的假设检验问题,建立了检验统计量,同时获得了检验统计量的极限分布.此外,我们也通过随机模拟的方法验证了方法的可行性.The problem of estimating and checking the overall parameters of the mixed exponential distribution with partial missing data is studied. On the one hand, the moment estimation of unknown parameters is given, at the same time, the coincidence and asymptotic normality of estimators are proved. On the other hand, we discussed the hypothesis testing that the two mixed exponential distributions are equal to each other, established the test statistics, at the same time, the limit distribution of test statistics is obtained. In addition, we also verified the feasibility of the method by means of stochastic simulation.

关 键 词：混合指数分布 缺失数据 矩估计 假设检验 

分 类 号：O212[理学—概率论与数理统计]