Hi,
You can find some references or algorithms in the following three books.
1. Causation, prediction, and search
Peter Spirtes, Clark Glymour, and Richard Scheines ; with additional
material by David Heckerman ... [et al.] Cambridge, Mass. : MIT Press, c2000. Edition 2nd ed.
2. Computation, causation, and discovery
edited by Clark Glymour and Gregory F. Cooper. Menlo Park, Calif. : AAAI Press ; Cambridge, Mass. : MIT Press, c1999
3. Causality : models, reasoning, and inference
Judea Pearl. Cambridge, U.K. ; New York : Cambridge University Press, 2000
In addition, a freely available implementation (an ASP solution) can be found at http://b-course.hiit.fi (also with some supporting material).
Go all the way through the "D-trail" (use the example data sets if you do not have data of your own), and then choose either "Naive causal model" or "Not-so-naive causal model". The first option uses models with no latent variables, while the second considers also models with latent variables that are parents of exactly two observed variables.
All the best,
--Petri
-- Petri Myllym�ki, Ph.D., Professor Department of Computer Science, University of Helsinki, Finland [EMAIL PROTECTED], www.cs.helsinki.fi/petri.myllymaki GSM: +358 40 553 1162, Office: +358 9 191 51337, Fax: +358 9 694 9768
_______________________________________________ uai mailing list [email protected] https://secure.engr.oregonstate.edu/mailman/listinfo/uai
