Bandwidth selection for a data sharpening estimator in nonparametric regression

被引:3
|
作者
Naito, Kanta [1 ]
Yoshizaki, Masahiro [1 ]
机构
[1] Shimane Univ, Dept Math, Matsue, Shimane 6908504, Japan
基金
日本学术振兴会;
关键词
Bandwidth; Bias reduction; Data sharpening; Kernel; Nonparametric regression; Plug-in method; LEAST-SQUARES REGRESSION; BIAS REDUCTION METHOD; DENSITY-ESTIMATION;
D O I
10.1016/j.jmva.2008.12.016
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
This paper is concerned with data-based selection of the bandwidth for a data sharpening estimator in nonparametric regression. Two kinds of bandwidths are considered: a bandwidth vector which has a different bandwidth for each covariate, and a scalar bandwidth that is common for all covariates. A plug-in method is developed and its theoretical performance is fully investigated. The proposed plug-in method works efficiently in our simulation study. (c) 2009 Elsevier Inc. All rights reserved.
引用
收藏
页码:1465 / 1486
页数:22
相关论文
共 50 条