区间删失数据情形下GE分布参数的最大似然估计(英文)  

The MLE of GE distribution with interval censored data

作　　者：程从华[1] 陈进源[2] 

机构地区：[1]湛江师范学院数学与计算科学学院,广东湛江524048 [2]兰州大学数学与统计学院,兰州730000

出　　处：《兰州大学学报（自然科学版）》2012年第3期113-117,共5页Journal of Lanzhou University(Natural Sciences)

基　　金：Supported by the National Natural Science Foundation of China(71171103);Specialized Research Fund for the Doctoral Program of Zhanjiang Normal University(ZL1110)

摘　　要：讨论了样本数据为区间型数据时参数的最大似然估计问题.当数据为区间删失情形时,参数最大似然估计的精确表达式不存在,甚至近似解都很难得到.由于区间型数据是一种不完全数据,所以利用EM算法来求参数的近似最大似然估计.为了演示,提供了一个真实寿命数据分析的实例.The two-parameter maximum likelihood estimation （MLE） problem for the GE distribution with congideration of interval data was studied. In the presence of interval data, the analytical forms for the restricted MLE of the parameters of GE distribution did not exist. Since the interval data were somewhat incomplete, we proposed to use EM algorithm to compute the MLEs of the parameters. A real life data analysis was presented to illustrate the method.

关 键 词：广义指数分布 EM算法 最大似然估计 区间数据 

分 类 号：O212.1[理学—概率论与数理统计] O213.2[理学—数学]