Exact asymptotic L1-error of a kernel density estimator under censored data

被引:3
|
作者
Lemdani, M
Ould-Saïd, E
机构
[1] Univ Lille 2, Fac Pharm, Lab Biomath, F-59006 Lille, France
[2] Univ Littoral Cote Opale, LMPAJ Liouville, F-62228 Calais, France
关键词
kernel estimator; censored data; L-1-error; strong approximation;
D O I
10.1016/S0167-7152(02)00278-X
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
In this paper, we give the exact asymptotic L-1-error for the kernel estimator of the density function from censored data. We also give asymptotically optimal bandwidths. Strong approximation of the product-limit estimator by a Gaussian process is used to obtain the result. (C) 2002 Elsevier Science B.V. All rights reserved.
引用
收藏
页码:59 / 68
页数:10
相关论文
共 50 条