In: Proceedings of the 19th Annual Symposium on Computer Applications in Medical Care (SCAMC95), 1995.
Clinical Simulation using ContextSensitive Temporal Probability Models
Peter Haddawy, Ph.D., James W. Helwig, B.S., Liem Ngo, M.S., Robert A. Krieger, M.S. Department of Electrical Engineering and Computer Science University of WisconsinMilwaukee, Milwaukee, Wisconsin 53201
We present a language for representing context sensitive temporal probabilistic knowledge. Con text constraints allow inference to be focused on only the relevant portions of the probabilistic knowledge. We provide a declarative semantics for our language and an implemented algorithm (BNG) that generates Bayesian networks to com pute the posterior probabilities of queries. We il lustrate the use of the BNG system by applying it to the problem of modeling the effects of medica tions and other interventions on the condition of a patient in cardiac arrest.