BOOTSTRAP SIMULTANEOUS ERROR BARS FOR NONPARAMETRIC REGRESSION

被引：185

作者：

HARDLE, W

MARRON, JS

机构：

[1] UNIV BONN,FAK RECHTS & STAATSWISSENSCH,W-5300 BONN 1,GERMANY

[2] CATHOLIC UNIV LOUVAIN,B-1348 LOUVAIN,BELGIUM

来源：

ANNALS OF STATISTICS | 1991年 / 19卷 / 02期

关键词：

BOOTSTRAP; ERROR BARS; KERNEL SMOOTHING; NONPARAMETRIC REGRESSION; VARIABILITY BOUND;

D O I：

10.1214/aos/1176348120

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：

Simultaneous error bars are constructed for nonparametric kernel estimates of regression functions. The method is based on the bootstrap, where resampling is done from a suitably estimated residual distribution. The error bars are seen to give asymptotically correct coverage probabilities uniformly over any number of gridpoints. Applications to an economic problem are given and comparison to both pointwise and Bonferroni-type bars is presented through a simulation study.

引用

页码：778 / 796

页数：19

共 25 条

[1]

[Anonymous], 1991, INTRO PROBABILITY TH

[2] SOME ASYMPTOTIC THEORY FOR THE BOOTSTRAP [J].