Learning Bayesian Networks: A Unification for Discrete and Gaussian Domains

We examine Bayesian methods for learn­ ing Bayesian networks from a combination of prior knowledge and statistical data. In particular, we unify the approaches we pre­ sented at last year's conference for discrete and Gaussian domains. We derive a gen­ eral Bayesian scoring metric, appropriate for both domains. We then use this metric in combination with well-known statistical facts about the Dirichlet and normal-Wishart dis­ tributions to derive our metrics for discrete and Gaussian domains.