Mining Positive and Negative Fuzzy Association Rules

While traditional algorithms concern positive associations between binary or quantitative attributes of databases, this paper focuses on mining both positive and negative fuzzy association rules. We show how, by a deliberate choice of fuzzy logic connectives, significantly increased expressivity is available at little extra cost. In particular, rule quality measures for negative rules can be computed without additional scans of the database.

• Type: Conference
• Year: 2004
• Where: KES
• Authors: Peng Yan, Guoqing Chen, Chris Cornelis, Martine De Cock, Etienne E. Kerre
• Keywords: Concern Positive Associations, Fuzzy Logic Connectives, KES 2004, Negative Fuzzy Association