Fuzzy Classifiers with a Two-Stage Reject Option

In general, fuzzy classifiers have high interpretability. They can linguistically explain the reason why an input pattern is classified as a particular class through the linguistic interpretation of each antecedent fuzzy set. A reject option that rejects patterns near the boundaries between different classes is an approach to increase the reliability of classifiers. However, the conventional threshold-based reject option may reject more patterns than necessary to achieve high reliability. In this paper, we propose a two-stage reject option where a machine learning model is used after the threshold-based decision by a fuzzy classifier. If the class labels predicted by the machine learning model and the fuzzy classifier are the same, the fuzzy classifier outputs the predicted class label without rejection. Through computational experiments, we discuss the trade-off relation between the accuracy and rejection rate by the proposed two-stage reject option using various machine learning models.