A0158
Title: A new proposal to mitigate multicollinearity in linear regression models
Authors: Aslam Muhammad - Bahauddin Zakariya University (Pakistan) [presenting]
Shakeel Ahmad - Bahauudin Zakariya University (Pakistan)
Abstract: A new general class of biased estimators is proposed which includes some popular estimators as special cases and discusses its properties for multiple linear regression models with the issue of multicollinearity. This proposal extends the existing Liu and Liu-type estimators. The performance of the proposed estimator is compared with many of the leading estimators, using the mean squared error matrix criterion. An extensive simulation study has shown an attractive performance of the proposed estimator. Finally, the application of the proposed estimator has been demonstrated with a chemical dataset.
