The Comparison of Bandwidth Selection Methods Using Kernel Function

This paper presents the bandwidth selection methods for local polynomial regression with Normal, Epanechnikov, and Uniform kernel function. The bandwidth selection methods are proposed by Histogram Bin Width method, Bandwidth for Kernel Density Estimation method, and Bandwidth for Local Linear Regression method to estimate the local polynomial regression estimator. Using Monte Carlo simulations, we compare the Mean Square Error (MSE) of the bandwidth selection methods. For simulation results, it can be seen that the MSE of Bandwidth for Kernel Density Estimation method provides the smallest in all situations. The bandwidth selection methods are applied to the Stock Exchange of Thailand (SET) index. The results show that the MSE of Kernel Density Estimation method with Normal kernel function is the smallest as the simulation study.

Keywords: Bandwidth, Epanechnikov kernel function, Local linear regression, Local polynomial regression

E-mail: kaautcha@kmitl.ac.th