Application of semi-parametric single index model on the quantitative data of white oil products

Abstract

   The Semi-Parametric Single Index Model has been studied and its parameters have been estimated by using The Weighted Average Derivative Estimation Method (WADE) , and The Link function has also been studied by using (Natarajan - Watson) Estimation .

   In the practical side, a practical application have been made on the data of the oil pipeline company and Building the semi-parametric single index model, The best estimate of the model is chosen by using several different types of Kernel functions based on comparative criteria  , The results showed that the best model in the data representation is the semi-parametric single index model (SSIM) in the Quartic function.

Key Word: Semi-Parametric Model, Semi-Parametric Single Index Model Weighted Average Derivative Estimation Method, (Natarajan – Watson) Estimator Kernel function, Bandwidth Parameter, Cross-Validation Criterion .