A Note on Posterior Probability Estimation for Classifiers

Abstract

One of the central themes in the classification task is the estimation of class posterior probability at a new point $\mathbf{x}$. The vast majority of classifiers output a score for $\mathbf{x}$, which is monotonically related to the posterior probability via an unknown relationship. There are many attempts in the literature to estimate this latter relationship. Here, we provide a way to estimate the posterior probability without resorting to using classification scores. Instead, we vary the prior probabilities of classes in order to derive the ratio of pdf's at point $\mathbf{x}$, which is directly used to determine class posterior probabilities. We consider here the binary classification problem.

8 pages