Graphical models are an intuitive visual way to represent a model consisting of a set of hypothesis and the connections between them. They are also known as belief networks, probabilistic networks, and Bayesian networks. A key advantage of graphical models is that they describe the local interconnections that collectively make the model work.
