An inferential measure of dependence between two systems using Bayesian model comparison
Abstract
We propose to quantify dependence between two systems and in a dataset based on the Bayesian comparison of two models: one, , of statistical independence and another one, , of dependence. In this framework, dependence between and in , denoted , is quantified as , the posterior probability for the model of dependence given , or any strictly increasing function thereof. It is therefore a measure of the evidence for dependence between and as modeled by and observed in . We review several statistical models and reconsider standard results in the light of as a measure of dependence. Using simulations, we focus on two specific issues: the effect of noise and the behavior of when has a parameter coding for the intensity of dependence. We then derive some general properties of , showing that it quantifies the information contained in in favor of versus . While some of these properties are typical of what is expected from a valid measure of dependence, others are novel and naturally appear as desired features for specific measures of dependence, which we call inferential. We finally put these results in perspective; in particular, we discuss the consequences of using the Bayesian framework as well as the similarities and differences between and mutual information.
Cite
@article{arxiv.2412.06478,
title = {An inferential measure of dependence between two systems using Bayesian model comparison},
author = {Guillaume Marrelec and Alain Giron},
journal= {arXiv preprint arXiv:2412.06478},
year = {2024}
}
Comments
To be published in IEEE Transaction on Systems, Man, and Cybernetics: Systems