Approach to generating rules for expert systems using rough set theory

Abstract

The problem of data mining and knowledge discovery as generating rules from databases has become a great research interest of many researchers. Many methods such as induction learning, ID3 etc., have been developed. A a new approach based on rough set theory has been proposed. Rough set theory was proposed by Zdzislaw Pawlak (1980) to deal with inconsistent problems. Our work is to apply this theory in extracting rules from given medical databases. This results in a set of decision rules, which will be provided for one of our diagnostic systems as a part of its knowledge base.