A novel approach to optimise fuzzy association rule by using evolutionary genetics algorithm

In data mining, extracting useful information from complicated and ambiguous datasets has remained a significant issue. To address such problems, the fuzzy logic approach is becoming increasingly very popular. The traditional association rule mining and clustering are two highly efficient approaches for finding underlying information in this discipline. This approach integrates the fuzzy logic theory to improve the efficiency of these two approaches. The integrated approach is more adaptive to dealing with real-world data. So, this study proposes a hybrid framework that effectively explores and optimises the number of generated rules. This work provides a unique cluster fuzzy association rules (CFARs) technique by combining fuzzy association rule mining and fuzzy C-means clustering. The proposed model generates three clusters for the fuzzy association rules for the considered 'online retail' dataset. This approach processes the primary set of CFAR to filter-out optimal number of rules. The optimisation process is carried out under the guidance of evolutionary genetics algorithm. The mining method extracts 5,000 fuzzy association rules from the considered data, which the optimisation process then reduces to 2,005 fuzzy association rules.