ppt课件-collective knowledge（集体知识）.pptVIP

Markov Networks Markov Networks Undirected graphical models Markov Networks Undirected graphical models Hammersley-Clifford Theorem If Distribution is strictly positive (P(x) 0) And Graph encodes conditional independences Then Distribution is product of potentials over cliques of graph Inverse is also true. (“Markov network = Gibbs distribution”) Markov Nets vs. Bayes Nets Inference in Markov Networks Goal: compute marginals conditionals of Exact inference is #P-complete Conditioning on Markov blanket is easy: Gibbs sampling exploits this Markov Chain Monte Carlo Gibbs Sampler