PyImpetus is a Markov Blanket based feature selection algorithm that produces the best possible combination of features by considering feature performance both individually as well as conditioned on other, already selected features.
