Memoirs of the Graduate School of Science and Engineering, Shimane University. Series B, Mathematics

number of downloads : ?
Memoirs of the Graduate School of Science and Engineering, Shimane University. Series B, Mathematics 44
2011-03 発行

Kernel regression for binary response data

Okumura Hidenori
File
Description
This paper is based on the author's thesis, \Kernel Regression for Binary Response Data. We consider kernel-based estimators with additional weights of the regression functions in nonparametric binomial and binary regressions. Firstly, in the binomial regression with a single covariate and a xed covariate design, we introduce a weighted Nadaraya-Watson estimator and its bias adjusted estimator discussed in Okumura and Naito [13]. Secondly, in the binary regression with multiple covariates and a random covariate design, we propose weighted local linear estimators.