A novel fuzzy rule-based classification system Based on Classifier Selection Strategy

摘要

Fuzzy systems have been used as a mechanism to build classifiers which are called fuzzy rule based classification systems (FRBCSs). In this paper, a new method for improving this kind of classifiers, based on ensemble strategy, is proposed. Here instead of building a classifier or a fusion of a group of them, we build some base classifiers and select one for every test pattern. A number of UCI datasets are used to assess the performance of the proposed method in comparison with reward and punishment and another method. Simulation results show our method's performance is a notch above these schemas.