BOOTSTRAPPING THE MEAN INTEGRATED SQUARED ERROR

被引：51

作者：

CAO, R ^{[1
]}

机构：

[1] UNIV LA CORUNA,E-15071 LA CORUNA,SPAIN

来源：

JOURNAL OF MULTIVARIATE ANALYSIS | 1993年 / 45卷 / 01期

关键词：

BANDWIDTH SELECTION; BOOTSTRAP; PILOT ESTIMATION;

D O I：

10.1006/jmva.1993.1030

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：

A smooth bootstrap method is used to find an estimator of the mean integrated squared error in density estimation. This provides a natural bootstrap selector for the bandwidth. Some rates of convergence and limit distributions are presented for this new selector as well as for some modification of it. © 1993 Academic Press Inc.

引用

页码：137 / 160

页数：24

共 23 条

[1]

BOWMAN AW, 1984, BIOMETRIKA, V71, P353

[2]

CAO R, 1993, IN PRESS COMP STATIS

[3]

CUEVAS A, 1991, NONPARAMETRIC FUNCTI, P225

[4]

FALK M, 1990, BOOTSTRAP OPTIMAL BA