Association Rule Mining with Dynamic Adaptive Support Thresholds for Associative Classification

The first and foremost task in any associative classification algorithm is mining of the association rules. Many studies have shown that the minimum support measure plays a key role in building an accurate classifier. Without the knowledge of the items and their frequency, user specified support measures are inappropriate, seldom may they coincide. .In this paper, we propose an approach called DASApriori i.e) Dynamic Adaptive Support Apriori to calculate the minimum support for mining class association rules and to build a simple and accurate classifier. Our experiments on 5 databases from UCI repository show that it achieves the best balance between the rule set size and classification accuracy even without the use of rule pruning techniques when compared with other associative classification approaches.