Improved Kibria-Lukman Type Estimator:Application and Simulation

The method of ordinary least square (OLS) in the linear regression model is widely used in different fields for quite some time, and it is grossly affected by multicollinearity. OLS estimator often exhibits unstable and unreliable results when there is multicollinearity. The ridge regression estimator had been widely accepted as a substitute to OLS estimator when there is multicollinearity. Recently, Kibria and Lukman (2020) developed the KL estimator and found it preferable to the ridge estimator. In this study, we modified the KL estimator to propose a new estimator. The new estimator is called the Modified KL estimator. Simulation study and real-life application were carried out to compare the performance of this new estimator and some other existing estimators. Theoretical comparison using data set showed MSE difference of 92.0896, 81.9774 and 338.9007 between the Modified KL estimator and the Ridge, Liu and KL estimators respectively. The simulation study and the real-life application results show that the proposed estimator consistently dominate other estimators considered in the study using the MSE as criterion.