A new estimator by generalized modified jackknife ridge regression estimator

Abstract EN

multicollinearity is an important problem in regression which produces undesirable effects on the least squares estimator.

Ridge regression is one of the well known solutions to the multicollinearity problem.

Specifically, the generalized ridge regression (GRR) estimation leads to reduction in the sampling variance whereas jackknifed ridge regression (JRR) estimation leads to reduction in the bias.

Batah et al.

[1] proposed modified jackknife ridge regression estimator (MJR) by combining the ideas of GRR and JRR estimators.

In this article, we propose a new estimator namely, generalized Jackknife ridge regression estimator (GJR) by generalizing the MJR.

We showed that new proposed estimator (GJR) is superior in the mean square error (MSE) sense to the LASSO estimator, generalized ridge regression estimator, jackknifed ridge regression estimator and modified jackknifed ridge regression estimator.