Finding relevant attributes and membership functions

Fuzzy systems that automatically derive fuzzy if-then rules from numeric da ta have been developed Most have to predefine membership functions in order to learn. Hong and Lee proposed a general learning method that automatical ly derives fuzzy if-then rules and membership functions from a set of given training examples using a decision table. All available attributes were in cluded in the decision table and the initial membership functions for each attribute were built according to the predefined smallest unit. Although Ho ng and Lee's method accurately derives the fuzzy if-then rules and final me mbership functions, the decision table and the initial membership functions are complex if there are many attributes or if the predefined unit is smal l. We improve Hong and Lee's method by first selecting relevant attributes and building appropriate initial membership functions. These attributes and membership functions are then used in a decision table to derive final fuz zy if-then rules and membership functions. Experimental results on Iris dat a show that the proposed method effectively induces membership functions an d fuzzy if-then rules. (C) 1999 Elsevier Science B.V. All rights reserved.