3 The Bayesian Network Representation 45 3.1 Exploiting Independence Properties 45 3.1.1 Independent Random Variables 45 3.1.2 The Conditional Parameterization 46 3.1.3 The Naive Bayes Model 48 3.2 Bayesian Networks 51 3.2.1 The Student Example Revisited 52 3.2.2 Basic Independencies in Bayesian Networks 56 3.2.3 Graphs and Distributions 60 3.3 Independencies in Graphs 68 3.3.1 D-separation 69 3.3.2 Soundness and Completeness 72 3.3.3 An Algorithm for d-Separation 74 3.3.4 I-Equivalence 76 3.4 From Distributions to Graphs 78 3.4.1 Minimal I-Maps 79 3.4.2 Perfect Maps 81 3.4.3 Finding Perfect Maps ⋆ 83 3.5 Summary 92 3.6 Relevant Literature 93 3.7 Exercises 96