Comparison of quantile regression and quantile neural networks

Regression analysis is a statistical modelling method used to examine the relationship between variables. In this method, the average of the response variable is tried to be estimated based on the values of the explanatory variables. Regression analysis is based on some assumptions. If these assumptions are met, the least squares method gives the best estimates. But real data may not always provide these assumptions. New methods have been developed for such cases where assumptions are not provided. The quantile regression is one of these methods. In this thesis, quantile regression and the use of quantile regression in artificial neural networks were examined and these models were compared on simulation data.