A METHOD FOR COMPUTING ALL MAXIMALLY GENERAL RULES IN ATTRIBUTE-VALUESYSTEMS

A method for finding all deterministic and maximally general rules for a target classification is explained in detail and illustrated with e xamples: Maximally general rules are rules with minimal numbers of con ditions. The method has been developed within the context of the rough sets model and is based on the concepts of a decision matrix and a de cision function. The problem of finding ail the rules is reduced to th e problem of computing prime implicants of a group of associated Boole an expressions. The method is particularly applicable to identifying a ll potentially interesting deterministic rules in a knowledge discover y system but can also be used to produce possible rules or nondetermin istic rules with decision probabilities, by adapting the method to the definitions of the variable precision rough sets model.