A Comparisons Among Some Biased Estimators in Generalized Linear Regression Model in present of Multicollinearity

Abstract

The Multicollinearity problem has currently became known by many researchers and knowledge of the statistical effects on parameters of the multiple linear regression model.In a simple case this problem causes to move away the estimate of parameters in the regression model that he scientific capabilities that desired in interpretation of the phenomenon in a correct way.This problem has been found in many areas that has been got negative effects on the estimates and variances of coefficients of (OLS). So we should avoid this problem and develop appropriate solution. In this article we will present some methods to estimate a (GRR, GJR, GL) to overcome this problem. The aim is to select the best estimator for the multiple linear regression model in case presence of Semi Perfect Multicollinearity among the explanatory variables by using Monte Carlo method. Then, We will compare among the estimators by using MSE. Finally, We conclude that (GL) is the best method.Keywords: Multicollinaerity ,Generalized Ridge Regression (GRR), Generalized Jackknife Ridge Regression (GJR), Generalized Liu Estimator (GL).