Using Markov Blankets for Causal Structure Learning

Abstract

We show how a generic feature-selection algorithm returning strongly relevant variables can be turned into a causal structure-learning algorithm. We prove this under the Faithfulness assumption for...