A computable theory for learning Bayesian networks based on MAP-MDL principles

- Pan, Heping, McMichael, Daniel