截断样本下的强率函数和密度函数核估计的渐进正态性(英文)  

Asymptotic Normality of Kernel Estimates for Intensity and Density Functions with Censored and Truncated Data

作　　者：冯琦琳 卢江 

机构地区：[1]北京成人教育学院,北京100101 [2]魁北克大学蒙特利尔分校

出　　处：《应用概率统计》1998年第2期122-130,共9页Chinese Journal of Applied Probability and Statistics

摘　　要：本文用[1]发展的计数过程去研究截断样本下强率函数核估计的渐进正态性．在弱于[7]和[10]的条件下，得到了更一般的结果．接着我们将这种方法运用到密度函数核估计，在较弱的条件下，得到了截断样本下密度函数核估计的渐进正态性．In this paper, we use the so called counting process, developed by Aalen [1], to study the asymptotic normality of kernel estimates for the intensity function based censored or truncated data. Our results are more generious, however the assumptions we make are weak comparing with those of Ramlau-Hansen [7] and Uzunogullri and Wang [10]. Subsequently, we use the counting process method to get the asymptotic normality of kernel estimates for the density function under mild conditions.

关 键 词：渐进正态性 密度函数 强率函数 核估计 

分 类 号：O212.7[理学—概率论与数理统计]