WHAT IS INSIDE A PROBABILISTIC EXPERT SYSTEM

Abstract

This paper is an outline of how a probabilistic expert system works, therefore the theoretical details have been omitted. They have been widely dealt with e. g. in the quoted references. The framework is the one of de Finetti’s theory. His fundamental theorem allows us to consider at first only the probability distribution on some of the events of interest and then, when new information gets into the system, to extend to other conditional events in a coherent way. Five examples that consider several typical situations one can come up against are illustrated. They allows us to have a general understanding of how a probabilistic expert system works.