Paper Title
Mixed Estimator of Kernel and Fourier Series in Nonparametric Regression
Abstract
Lets paired the observation follow the additive nonparametric regression model where
and Random errors is a normal distribution with mean 0 and variance . The aim of this study is obtain a mixed estimator In order to accomplish the aim, the regression curve is approached by kernel with bandwidths and the regression curve component is fourier series where it is approached by with oscillation paremeter M. The estimator is where Based on Penalized Least Squares (PLS) method
with smoothing parameter , the estimator is and So that, the mixed estimator is
where Matrix and are depended on bandwidths parameter , smoothing parameter and oscillation paremeter M. Optimal , and M can be obtained by using Generalized Cross Validation (GCV).
Keywords- Mixed Nonparametric Regression, Kernel, Fourier Series