What is the relation between message passing and probabilities in Bayesian inference?

Asked May 14 '19 at 21:34

Active May 14 '19 at 21:34

Viewed 400 times

The belief propagation algorithm is a message passing algorithm that can be used to estimate marginal probabilities on Bayesian networks.

What is the definition of these messages? What is the relation between the messages and the marginal probability of a random variable? How come that we can retrieve marginal probabilities from an iterative procedure that passes messages between nodes?

asked May 14 '19 at 21:34

Message passing is a way to eliminate variables. Please refer to this note: https://ermongroup.github.io/cs228-notes/inference/jt/ – Lerner Zhang May 14 '19 at 22:12
Then I thought this would be of help: https://ermongroup.github.io/cs228-notes/inference/ve/ – Lerner Zhang May 14 '19 at 22:57

What is the relation between message passing and probabilities in Bayesian inference?

0 Answers0