Structure learning and parameter estimation of discrete Bayesian networks using likelihood-based criteria. Exhaustive search for fixed node orders and stochastic search of optimal orders via simulated annealing algorithm are implemented.