The Probabilistic and Geometric Approach to Multivariate Logistic Regression

This paper gives an overview of the two approaches to implement the Multivariate Logistic Regression. There is the Geometric way and the Probabilistic way. The aim of this paper is to establish that both the approaches effectively produce the same result and the Probabilistic approach can be converted to the geometric approach with the help of some mathematical manipulation. The paper is to prove mathematically that the geometric approach which divides the classes using hyperplane and the Probabilistic method that uses the log of odds ratio to divide the classes are essentially the same and the Probabilistic approach with some mathematical manipulation can essentially be converted to the equation of hyperplane.