Message-passing method
From Wikipedia, the free encyclopedia
Message-passing methods are a set of algorithms in statistics/machine learning for doing inference through local computation. Belief propagation on Bayesian networks is a good example of a message-passing method.