Methods of Obtaining the Ridge Parameter K in Multiple Linear Regression Analysis

Multiple Linear Regression is a form of analysis or technique that shows how a continuous dependent variable is related to two or more independent variables. It plays a significant role in the general linear model. When two or more predictor variables are associated, a challenge is always encountered in the use of Multiple Linear Regression. This challenge is called Multicollinearity that can inflate the Least Squares Regression Coefficient estimates, which are dependent on the correlated predictor variables in the model. Researchers in several fields such as Finance and Economics have been faced with this unavoidable difficulty. Multicollinearity can cause inaccurate coefficient variances and unstable estimates, making it harder to decide the fit model. Hence, to address multicollinearity, alternative modalities like Ridge Regression have been offered. The value of the Ridge Parameter k, as scalar, vector, and matrix, can be calculated in different ways. This paper tries to determine the optimal Ridge Parameter as per the mean squares error criterion.