删失样本及完全样本下含位置参数的多元指数分布的参数估计  

Parameter Estimation for the Multivariate Exponential Distribution Which Has a Location Parameter Under Censored Samples or Complete Samples

作　　者：李国安[1] 李建峰[1] 王立洪[1] 

机构地区：[1]宁波大学金融工程系,宁波315211

出　　处：《系统科学与数学》2017年第8期1854-1865,共12页Journal of Systems Science and Mathematical Sciences

基　　金：宁波大学学科项目(XKL14D2037)资助课题

摘　　要：首先归纳出指数分布抽样基本定理;通过分布的识别性分析及参数的识别性分析,导出了串联系统下多元Marshall-Olkin型指数分布的一个识别特征,并由此得到了基于删失样本的含位置参数的多元Marshall-Olkin型指数分布参数的最大似然估计及一致最小方差无偏估计;通过密度分拆重组技术,还导出了随机系统下多元Marshall-Olkin型指数分布的一个特征,并由此得到了基于完全样本的含位置参数的二元Marshall-Olkin型指数分布参数的最大似然估计及无偏估计.The fundamental sampling theorem of exponential distribution is con- cluded in this paper. Through identifiability analysis of the distributions and analysis of parameter identification, an identifiable characteristic of multivariate Marshall- Olkin exponential distribution is derived based on series system, thus, the maximum likelihood estimator and uniformly minimum variance unbiased estimator of the multivariate Marshall-Olkin exponential distribution with location parameter under censored samples are obtained. Through the analysis of partition and reorganization of density, also a characteristic of multivariate Marshall-Olkin exponential distribu- tion is derived based on random system, hence, the maximum likelihood estimator and unbiased estimator of the bivariate Marshall-Olkin exponential distribution with location parameter under complete samples are obtained.

关 键 词：指数分布抽样基本定理 含位置参数的多元指数分布 特征 最大似然估计 一致最小方差无偏估计 删失样本 完全样本 

分 类 号：O212.1[理学—概率论与数理统计]