Fig. 1From: Discovering causal interactions using Bayesian network scoring and information gainOn the left is a Bayesian network representing a causal interaction with no marginal effects, and on the right is a Bayesian network representing a causal interaction described by the Noisy-Or modelBack to article page