A NEW MODIFICATION TO THE ADAPTIVE NADARAYA-WATSON KERNEL REGRESSION ESTIMATOR

被引:1
|
作者
Joshi, Venkatesh B. [1 ]
Deshpande, Bhargavi [1 ]
机构
[1] Karnatak Univ, Dept Studies Stat, Dharwad 580003, Karnataka, India
关键词
nonparametric regression; Nadaraya-Watson estimator; adaptive Nadaraya-Watson estimator; modified Nadaraya-Watson estimator; mean squared error;
D O I
10.17654/AS049040245
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, a modification to the adaptive Nadaraya-Watson (NW) kernel regression estimator based on a function of the range of pilot density estimates and the sample size with varying bandwidths is proposed. The performance of proposed estimator in terms of mean squared error (MSE) is compared with NW and modified NW estimators. The results of a simulation study reveal that the proposed estimator provides better estimates in interpolating data with a minimum rate of convergence of 10(-6).
引用
收藏
页码:245 / 256
页数:12
相关论文
共 50 条