부산대학교 도서관

The association rules have been demonstrated to beuseful for classification, like CBA, CMAR. However, the rule mining procedure does not focus on classification, and most of the rues will be pruned in classification. As a result, if we could directly mine the classification as sociation rules, the time will be saved. Mean while, we could expect even better accuracy because the mining procedure it self considers the classification. In this paper, we build a novel evolution ary method, named Evo CMAR, to directly mine classification association rules.