Factorization of Discrete Probability Distributions
Dan Geiger, Christopher Meek, Bernd Sturmfels · arXiv (Cornell University) · 2012
We formulate necessary and sufficient conditions for an arbitrary discrete probability distribution to factor according to an undirected graphical model, or a log-linear model, or other more general exponential models. This result generalizes the well known Hammersley-Clifford Theorem.