指数分布均值估计量的比较  

Comparison of Mean Estimates of Exponential Distribution

作　　者：蔡明书 李树有[1] 宓颖[1] CAI Ming-shu;LI Shu-you;MI Ying(College of Science,Liaoning University of Technology,Jinzhou 121001,China)

机构地区：[1]辽宁工业大学理学院

出　　处：《辽宁工业大学学报（自然科学版）》2019年第5期288-294,共7页Journal of Liaoning University of Technology(Natural Science Edition)

摘　　要：对于两个给定的指数分布总体,基于截尾样本,讨论其均值θi(i=1,2)极大似然估计的比较问题。分别就定时截尾样本和定数截尾样本,给出θi(i=1,2)极大似然估计满足随机序的充分条件。最后,针对双参数指数分布,基于逐次结尾样本,给出参数极大似然估计满足随机序的充分条件,并加以证明。For two given exponential distribution populations, the comparison problem of the maximum likelihood estimation of the mean θi(i=1,2) is discussed based on the truncated samples. The timed censored samples and the censored samples are given respectively, and the sufficient conditions for the θi(i=1,2) maximum likelihood estimation to satisfy the random order are given. Finally, for the two-parameter exponential distribution, based on the successively ending samples, the sufficient conditions for the parameter maximum likelihood estimation to satisfy the random order are given and proved.

关 键 词：指数分布 极大似然估计(MLE) Ⅰ型截尾样本 Ⅱ型截尾样本 Ⅱ型逐次截尾样本 

分 类 号：O212[理学—概率论与数理统计]