THE PREDICTION OF THE TWO PARAMETER RIDGE ESTIMATOR

Abstract

The prediction of regression model can be adversely affected by multicollinearity. Althoughbiased estimation procedures have been proposed as an alternative to least squares, there has been little analysisof the predictive performance of the resulting equations. Therefore, we discuss the predictive performanceof the Two Parameter Ridge (2PR) estimator compared to ordinary least squares, principal components andridge regression estimators. Also, the theoretical results are illustrated by numerical example and region is established where the 2PR estimator is uniformly superior to the other estimators.