On Restricted Shrinkage Jackknife Biased Estimator for Restricted Linear Regression Model

Abstract

In restricted linear regression model, more methods proposed to address the Multicollinearity problem and the high variance. For example, shrinkage biased estimation and optimization (Lagrange function). In this paper, we propose new biased estimator based on philosophy of Jackknife with the restricted least squares estimator. A new estimator called Restricted Shrinkage Jackknife estimator (RSJ). Also, we show that the statistical properties of new estimator with some theorems to compare the performance of new estimator with some restricted estimators and we make simulation study of these estimators. Finally, a real data has been taken into consideration to demonstrate how well the estimators perform.

Keywords

• restricted regression model
• jackknifed biased estimator
• multicollinearity problem
• two parameters estimator