Fitting Characteristics of the Two-Parameter Estimator in Linear Regression

Abstract

- introduced the twoparameter (TP) estimator for the vector of parameters in a linear regression model. In this paper, several better fitting characteristics of the TP estimator than those of the ordinary least squares and ridge estimators are considered. Furthermore, we propose and investigate the prediction error sum of squares (PRESS) statistics for selecting the parameter d of the TP estimator.