Nonparametric Kernel Estimation and Its Practical Application

This paper deals with the problem of nonparametric kernel estimation, particularly nonparametric kernel estimation of regression function in real life situations. There are many fields of application (Gross Domestic Product, Unemployment, Stock Market, etc.) and therefore this paper is focused on estimation of regression function of the selected underlying asset. First, there is described nonparametric regression including the influences of the main parameters (smoothing parameter, kernel function, etc.) on the shape of regression function. Then, there is analyzed and described in detail smoothing parameter and its estimation in relation to the regression curve. The obtained results are applied to generated data collection and real data collection of underlying asset (CEZ).