2015•Indian Journal of Science and TechnologyOpen access
Simulation Study to Verify the Appropriate k Value for Ridge Regression in Two-variable Regression Model
Hanan Duzan, Nurul Sima Binti Mohamad Shariff
Open full text 4 citations
Abstract
This study investigates the problem of using Ordinarily Least Squares (OLS) estimators in the presence of multicollinearity in regression analysis. As an alternative of OLS is ridge regression, which it is believed to be superior to least-squares regression in the presence of multicollinearity. The robustness of this method is investigated and comparison is made with the least squares method via simulation studies. Our results have shown that the system stabilizes in a region of k, which k is a positive quantity less than one and whose values depend on the degree of correlation between the independent variables. The results illustrate that k is a non-linear function of the correlation between the independent variables (r12). Keywords: Least Squares Method, Linear Models, Multicollinearity, Ridge Regression