Recursive calculation of probability distributions for sequentialdecision analysis models

摘要

Presents a general-purpose procedure to compactly specifynprobability information and efficiently calculate probabilityndistributions over the outcome evaluation measure for a sequentialndecision analysis model. Computer code to implement the calculationnprocedure is presented in the Pascal programming language. Empiricalnresults are reviewed that show that this approach can quickly determinenoutcome probability distributions for large sequential decision analysisnmodels using modest computing resources