Output details
11 - Computer Science and Informatics
Queen Mary University of London
Optimizing the Calculation of Conditional Probability Tables in Hybrid Bayesian Networks Using Binary Factorization
<13>This work extends the applicability of Bayesian Networks with both discrete and continuous variables by developing a new method for optimizing the calculation of conditional probability tables (CPTs) involving continuous nodes, using binary factorization of the arithmetical expressions declared to generate the CPTs for continuous nodes for deterministic functions and statistical distributions. This algorithm has now been included in the commercial BN package, AgenaRisk, for release alongside our major new textbook in the area. This work is also proving a crucial foundation for our current research into actuarial modelling of financial portfolios of causally dependent assets.