A two sample test based on mutual information

Abstract

For two independently drawn samples from continuous distributions, a permutation test based on the mutual information statistic is proposed for the null hypothesis that both the samples originate from the same population. It is demonstrated through simulation that this test is more powerful than the commonly used nonparametric tests in many situations. We also discuss the test in a real-life setup. Keywords and phrases: Nonparametric Test; Two-Sample Test; Information Theory; Permutation Test; Kernel Density Estimation.