SOME IMPROVED CLASSIFICATION-BASED RIDGE PARAMETER OF HOERL AND KENNARD ESTIMATION TECHNIQUES

Authors : Adewale F LUKMAN, Kayode AYİNDE

Pages : 93-106

View : 17 | Download : 8

Publication Date : 2016-09-01

Article Type : Research Paper

Abstract :In a linear regression model, it is often assumed that the explanatory variables are independent.This assumption is often violated and Ridge Regression estimator introduced by Hoerl and Kennard insert ignore into journalissuearticles values(1970);has been identified to be more efficient than ordinary least square insert ignore into journalissuearticles values(OLS); in handling it. However, it requiresa ridge parameter, K, of which many have been proposed. In this study, estimators based on Hoerl andKennard were classified into different forms and various types and some modifications were proposed toimprove it. Investigation were done by conducting 1000 Monte-Carlo experiments under five insert ignore into journalissuearticles values(5); levels ofmulticollinearity, three insert ignore into journalissuearticles values(3); levels of error variance and five levels of sample size. For the purpose of comparingthe performance of the improved ridge parameter with the existing ones, the number of times the MSE ofthe improved ridge parameter is less than the existing ones is counted over the levels of multicollinearity insert ignore into journalissuearticles values(5);and error variance insert ignore into journalissuearticles values(3);. Also, a maximum of fifteen insert ignore into journalissuearticles values(15); counts is expected. Results show that the improvedridge parameters proposed in this study are better than the existing ones most especially with the quantity
Keywords :