An intuitive markov networks (MRFs) tutarial?

Question

I would like to know about basics of markov networks (MRFs). Does anyone know an intuitive tutorial about the subject. I just need to very basics information about undirected graphical models.For example how to compute the joint distribution and conditional probabilities of a simple undirected graph(e.g A-B-C)?

Answer 1

Actually your example is a bit special. In that case I would implement the "Belief Propagation". In general it gives an approximate solution. But in your case it is exact. You can find a, imho, very good tutorial at: https://class.coursera.org/pgm/lecture/49

A complete lecture on Graphical Models is provided in the lecture https://class.coursera.org/pgm/