Abstract

The non-Parametric models difference from non-parametric statistic . It does not depend on determination or prior-hypothesis in building the model structure ,but it depends directly and basically on data .The nature and the number of the parameters is flexible and not fixed. Therefore ,these models are called free distribution models. The methods of non-parametric inference are mathematical processes in order to test statistical hypothesis which do not impose the existence of assumption about frequent distributions of variables, so they are less strength than parametric tests, but they are more robust in case of trespass the basic assumption or the way in which they can not be verified . In this study we estimate the non-parametric regression function directly without depening on any certain parameters by using Nadaraya-Watson methods by using the simulation