Dear Colleagues,
This note concerns learning Bayesian networks from data, anarea in which
I wrote a book in 2003. However, since then I have not kept that close
of a track of developments in the area. The GES algorithm assumes the
composition property, andthe constraint-based PC algorithm andmore
advanced constraint-based algorithms assume faithfulness or embedded
faithfulness. So none of them would discovera DAG in which two variables
together have an effect on a third variable, but neither of the
variables has a marginal effect. I am wondering if there are any
heuristic searchalgorithms, in a particular ones implemented in
available software, that address this situation. Clearly, there are
modifcations of these algorithms that would do so.
Thanks,
Rich
--
Richard E. Neapolitan, Ph.D., Professor
Division of Health and Biomedical Informatics
Department of Preventive Medicine
Northwestern University Feinberg School of Medicine
750 N. Lake Shore Drive, 11th floor
Chicago IL 60611
_______________________________________________
uai mailing list
uai@ENGR.ORST.EDU
https://secure.engr.oregonstate.edu/mailman/listinfo/uai
