Fuzzy If-Then Rules Classifier on Ensemble Data

Abstract

This paper introduces a novel framework in which classification method based on fuzzy IF-THEN rules is co-operated with an ensemble system. This model tackles several drawbacks. First, IF-THEN rules approaches have problem with high dimensional data since computational cost is exponential function. In this framework, rules are operated on outputs of base classifiers which have frequently lower dimension than original data. Moreover, outputs of base classifiers are scaled within the range [0, 1] so it is convenient to apply fuzzy rules directly instead of data transformation and normalization as requirement for inputs of fuzzy rules. The performance of this model was evaluated through experiments on 6 popular datasets from UCI Machine Learning Repository and comparisons with other state-of-art combining classifiers algorithms as well as other fuzzy IF-THEN rules approaches. The results show that our framework can improve the classification accuracy.