On Mining Fuzzy Classification Rules for Imbalanced Data

Fuzzy rule-based classification system (FRBCS) is a popular machine learningtechnique for classification purposes. One of the major issues when applying it onimbalanced data sets is its biased to the majority class, such that, it performs poorlyin respect to the minority class. However many cases the minority classes are moreimportant than the majority ones. In this paper, we have extended the basic FRBCSin order to decrease the side effects of imbalanced data by employing data-miningcriteria such as confidence and support. These measures are computed frominformation derived from data in the sub-spaces of each fuzzy rule. Theexperimental results show that the proposed method can improve the classificationaccuracy when applied on benchmark data sets.

کلیدواژه ها:

Imbalanced data-sets ، Fuzzy rule based classification systems ، Data-mining