Dear Sirs, I would appreciate if you could answer my question or reference me on some literature concerning the following.
I have learned a Bayesian network on 15 nodes with 120 data with C2 algorithm of Cooper and Herskovitz. What statistical tests could be done to prove the model? Since I have very few data, how can I test what probabilistic dependencies in my model are significant? Are there some investigations, how many data samples is needed to learn the dependencies between e.g. 15 variables? Thank you very much for your reply, Svetlana Bulashevska, Ph.D. Student.
