Abstract
The multiple linear regression model of the important regression models used in the analysis for different fields of science Such as business, economics, medicine and social sciences high in data has undesirable effects on analysis results . The multicollinearity is a major problem in multiple linear regression. In its simplest state, it leads to the departure of the model parameter that is capable of its scientific properties, Also there is an important problem in regression analysis is the presence of high leverage points in the data have undesirable effects on the results of the analysis , In this research , we present some of the robust methods in the multiple linear regression model These methods include the (Jackknife Ridge regression) methods based on the (MM) estimator and the (GM2) estimator (Modified Generalized M-estimator) . Using the Monte Carlo simulation, the two methods were compared in accordance with the comparison criterion, the mean squares error (MSE) and sample sizes (n = 20, n = 50, n = 100) and different pollution ratios (τ = 5%, 15%) , The comparison shows that (RJGM2) is the best method for estimating the parameters of the multiple linear regression model, which has the lowest value for mean squares error (MSE) compared with the rest of the other estimations.