Peter Spirtes
Department of Philosophy
Carnegie Mellon University
5000 Forbes Ave.
Pittsburgh, PA, 15213
E-mail: ps7z@andrew.cmu.edu
Phone: 412-268-8460
Fax: 412-268-1440
Greg Cooper
Center for Biomedical Informatics
University of Pittsburgh
Pittsburgh, PA 15213
E-mail: heckerma@microsoft.com
Phone: 412-647-7113
We tested a causal discovery algorithm on a database of pneumonia patients. The output of the causal discovery algorithm is a list of statements "A causes B", where A and B are variables in the database, and a score indicating the degree of confidence in the statement. We compared the output of the algorithm with the opinions of physicians about whether A caused B or not. We found that the doctors opinions were independent of the output of the algorithm. However, an examination of the output of results suggested a simple, well motivated modification of the algorithm which would bring the output of the algorithm into high agreement with the physicians opinions.
Bayesian networks, causal networks, causal discovery, medical informatics