The effect of partitioning for the performance of fuzzy associative classifiers

Abstract

Classification is one of the most popular and extensively applied techniques in data mining. The efficiency of a classification model is evaluated by two parameters, namely the accuracy and interpretability of the model. This paper proposes a fuzzy association rule-based classifier methodology that meets both criteria. Using the fuzzy concept, the obtained model is easily understandable and interpretable for the users. Since the accuracy of a classification model can be largely affected by the partitioning of numerical attributes, this paper discusses several fuzzy and crisp partitioning techniques. The effect of partitioning methods is examined on different case studies. The results of analysis show that classifier methods with fuzzy clustering based partitioning serve higher classification performance.