Comparison Some Robust Estimators for Estimate parameters logistic regression model to Binary Response – using simulation)).

Abstract

 The logistic regression model of the most important regression models a non-linear which aim getting estimators have a high of efficiency, taking character more advanced in the process of statistical analysis for being a models appropriate form of Binary Data.                                                          

Among the problems that appear as a result of the use of some statistical methods Is not to achieve some or all the requirements including the presence of abnormal values between data, appears when the data of the studied phenomenon are contaminated ,it means some of the observations variety clearly from other observations called outliers.                                                                       

From this point was the goal of this research to estimate parameters of logistic regression model through study some of Robust estimation methods The representing of the Robust weighted maximum likelihood estimators(WMLE), Quadratic Distance Estimators(QDE)  We Use Simulation  to comparison between two methods for different sample sizes and for  difference proportions of contamination through mean square error (MSE) of the model,  to reach the best method to estimate the parameter.                                                                      

It was Concluded in through this Research to advantage of the method (WMLE()) in estimate parameters of binary response logistic regression model for different of samples sizes.