I am doing researches about EM algorithm on learning Bayesian Network. To start with, I want to implement it by myself.
To implement EM, I need to estimate one posterior joint probability which is P(x, parents(x) | e) where x is a node and e is some evidence. Is there any way I can get this from smile? It seems pretty easy to get P(x | e), I can set some evidence and update the believe.
Thanks in advance.
1 post • Page 1 of 1