Abstract—Associative classification is a combination of association rule mining and classification for prediction. For mining associative classification, traditional algorithms generate the complete set of association rules, and then use a minimum confidence threshold to select interesting rules for classification. If the number of association rules is very large, it is time consuming to select only the interesting rules. In this paper, a new algorithm, called TOPAC (Top Associative Classification), is proposed to solve the problem. The TOPAC algorithm directly produces the interesting rules without the generation of candidate rules. Moreover, it discovers the interesting rules based on frequent closed itemsets to reduce the redundancy rules.
Index Terms—Associative classification, association rule, closed itemset, high confidence.
Panida Songram is with Department of Computer Science, Faculty of Informatics, Mahasarakham University, Mahasarakham, 44150, Thailand.
Cite: Panida Songram, "Mining Associative Classification without Candidate Rules," International Journal of Future Computer and Communication vol. 1, no. 2, pp. 138-141, 2012.
Copyright © 2008-2024. International Journal of Future Computer and Communication. All rights reserved.
E-mail: ijfcc@ejournal.net