Skip to Main Content
In this paper, a new data mining algorithm is proposed to enhance the capability of exploring interesting knowledge from databases with continuous values. The algorithm integrates Fuzzy Set Theory and ldquoGenetic Network Programming (GNP)rdquo to find interesting fuzzy association rules from given transaction data. GNP is a novel evolutionary optimization technique, which uses directed graph structures as gene instead of strings (Genetic Algorithms) or trees (Genetic Programming), contributing to creating quite compact programs and implicitly memorizing past action sequences. We adopt the Fuzzy Set Theory to mine associate rules that can be expressed in linguistic terms, which are more natural and understandable for human beings. The proposed method can measure the significance of the extracted association rules using support, confidence and chi2 test, and obtains a sufficient number of important association rules in a short time. Experiments conducted on real world databases are also made to verify the performances of the proposed method.