Factorization of Discrete Probability Distributions
arXiv:1301.0568
Abstract
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.
Appears in Proceedings of the Eighteenth Conference on Uncertainty in Artificial Intelligence (UAI2002)