删失数据下半变系数模型估计的渐近正态性  

The Asymptotic Normality of Estimation on Semivarying Coefficient Models with Censored Data

作　　者：吕士钦[1] 张日权[2,1] 卢准炜[1] 

机构地区：[1]太原理工大学理学院,太原030024 [2]山西大同大学数学系,大同037000

出　　处：《应用概率统计》2009年第6期641-648,共8页Chinese Journal of Applied Probability and Statistics

基　　金：山西省自然科学基金(2006011006)资助

摘　　要：半变系数回归模型是变系数回归模型的有效推广,已获得了广泛的应用.本文在响应变量随机删失下对其进行讨论,给出常系数和函数系数的估计,并证明了该估计的渐近正态性.Semivarying coefficient models are generalization of the varying-coefficient regression models, which have extremely wide applications. We studied them with censored data in this paper. By transform data, the estimator of parametric component and nonparametric component on semivarying coefficient models are developed. Asymptotic normalities of the estimator are investigated.

关 键 词：删失数据 半变系数模型 渐近正态性 

分 类 号：O212.7[理学—概率论与数理统计]