CMStatistics 2018: Start Registration
View Submission - CMStatistics
B1200
Title: A new modified Liu-type estimator for linear regression models with correlated regressors Authors:  Aslam Muhammad - Bahauddin Zakariya University (Pakistan) [presenting]
Abstract: A new biased estimator is presented and its properties for linear regression models when regressors are correlated are discussed. This new estimator is a general class of biased estimators which includes some well-known biased estimators as a special case. Comparison in the sense of mean squared error matrix (MSEM) is made with popular estimators i.e., the ordinary least squares estimator, ordinary ridge regression estimator, Liu estimator (LE) and Liu-Type estimators. It has been shown that our proposed estimator outperforms the other estimators. For the empirical study, the Monte Carlo simulations are made which show some superior performance of the new proposed estimators. An illustrative example has also been provided.