Mutual information is symmetric

Statement

Lemma

Suppose we have two random variables and over different domains and . We have that Mutual information is symmetric

Proof

This follows from Bayes rule, rules of logarithms and marginalisation.