Parameter-Estimation

learning bayesian networks from data — maximum-likelihood estimation of cpts, additive smoothing, handling missing data with expectation-maximisation, and structure learning from chow-liu trees to score-based search.