Classification Based on Consistent Itemset Rules

Abstract. We propose an approach to build a classifier composing consistent (100% confident) rules. Recently, associative classifiers that utilize association rules have been widely studied, and it has been shown that the associative classifiers often outperform traditional classifiers. In this case, it is important to collect high-quality (association) rules. Many algorithms find only rules with high support values, because reducing the minimum support to be satisfied is computationally demanding. However, it may be effective to collect rules with low support values but high confidence values. Therefore, we propose an algorithm that produces a wide variety of 100% confident rules including low support rules. To achieve this goal, we adopt a specific-to-general rule searching strategy, in contrast to previous approaches. Our experimental results show that the proposed method achieves higher accuracies in several datasets taken from UCI machine learning repository.

• Type: Journal
• Year: 2008
• Where: MLDM
• Authors: Yohji Shidara, Mineichi Kudo, Atsuyoshi Nakamura
• Keywords: Associative Classifiers, Low Support, Machine Learning, MLDM 2008, Support Values