相关G-M模型及其在不完全样本中的应用  

A Model of Correlated G-M and Applications in Non-completed Samples

作　　者：王勤[1] 

机构地区：[1]浙江工业大学理学院,浙江杭州310032

出　　处：《金华职业技术学院学报》2004年第2期48-50,52,共4页Journal of Jinhua Polytechnic

摘　　要：研究了相关情况下第i个顺序统计量及其标准化变量的期望、方差、协方差以及在各阶矩上的变换作用,将Gauss-Markov定理推广至一般相关的不完全样本的情形,并用最小二乘法对顺序统计量的参数作出了最佳线性无偏估计,推导出在样本值之间不相关或相关条件下计算估计量方差的公式。This paper investigates in the ith statistics under certain correlated circumstances.The ex-pectation,variance and covariance of its standardized variable as well as the transformation operation on moments are also studied in this paper.Based on the generalization of Gauss-Markov theorem to com-monly correlated non-completed samples,and the best linear unbiased estimators on the parameters of or-dered statistics using the least square method,the formula of computing variance on correlated or uncor-related samples is deduced.

关 键 词：BLUE 完成样本(数) 方差 顺序统计量 

分 类 号：O212[理学—概率论与数理统计]