Introduction

“Graphical models are a marriage between probability theory and graph theory” (Michael Jordan, 1999 [154]). The basic idea underlying probabilistic graphical models is to offer “a mechanism for exploiting structure in complex distributions to describe them compactly, and in a way that allows them to be constructed and utilized effectively” (Daphne Koller and Nir Friedman, 2009 [174]). They “have their origin in several scientific areas”, and “their fundamental and universal applicability is due to a number of factors” (Steffen Lauritzen, 1996 [189]). For example, the generality of graphical models is due to the fact that they “reveal the interrelationships between multiple variables and features of the underlying conditional independence” (Joe Whittaker, 1990 [315]). Moreover, “[c]omplex computations, required to perform inference and learning in sophisticated models, can be expressed in terms of graphical manipulations, in which underlying mathematical expressions are carried along implicitly” (Christopher Bishop, 2006 [30]).