摘要
在元件寿命服从双参数混合指数分布、元件个数服从几何分布的情形下,本文定义了混合指数几何分布,并研究了该分布的各种性质,讨论了参数的极大似然估计,并使用EM算法得到参数的近似估计。最后给出了参数的渐近方差、协方差、置信区间。
This paper defines the mixture exponential geometric distribution and analyzes its various properties based on the situation of the lifetime of the component and the number of the component subjects to two parameters mixture exponential distribution and geometric distribution,respectively. Furthermore,MLE of parameters is discussed and the approximate estimates of parameters are obtained by EM algorithm. Finally, this paper presents the parameters of asymptotic variances,covariances and confidence intervals.
出处
《统计研究》
CSSCI
北大核心
2015年第7期106-112,共7页
Statistical Research
基金
福建省自然科学基金“复杂网络度分布理论与应用研究”(2013J01004)资助
关键词
混合指数分布
几何分布
极大似然估计
EM算法
寿命分布
Mixture Exponential Distribution
Geometric Distribution
Maximum Likelihood Estimation
EM Algorithm
Lifetime Distribution