4 - Undirected probabilistic graphical models

Abstract

This chapter formally covers the undirected probabilistic graphical models, i.e., the Markov Networks (MNs). It starts with a definition of MNs, their properties, and different types of MNs. It then discusses various methods for performing exact and approximate inferences for MNs. This chapter continues to cover different methods for learning both MN structure and parameters. This chapter ends with a discussion of the commonalities and differences between BNs and MNs.