The choice of ply operator in fuzzy intelligent systems (Q582138)

Wen fuzzy modus ponens is used in a fuzzy intelligent reasoning system, the choice of ply (implication) operator is crucial. Different operators may endow the system with different characteristics, especially in situations with great uncertainty. The system may not provide the required expertise with some of the operators. Eleven operators have been classified into three categories and their properties studied. Modus ponens generating functions provide the value for the consequent. A minimal set of desired properties for the ply operator are detailed. Theoretical performance is compared with practical performance in the Fess system. Some comments on results from a related empirical study are provided. A framework for choosing a ply operator is given with justifications.