Catnet is a R package for structure learning and parameter estimation of discrete Bayesian networks using likelihood-based criteria.
It implements a dynamic programming algorithm for exhaustive search across networks with a given node order and stochastic search
of optimal node orders via simulated annealing algorithm.