FUZZY CLASSIFICATIONS USING FUZZY INFERENCE NETWORKS

In this paper, fuzzy inference models for pattern classifications have been developed and fuzzy inference networks based on these models are proposed. Most of the existing fuzzy rule-based systems have difficul ties in deriving inference rules and membership functions directly fro m training data, Rules and membership functions are obtained from expe rts. Some approaches use backpropagation (BP) type learning algorithms to learn the parameters of membership functions from training data. H owever, BP algorithms take a long time to converge and they require an advanced setting of the number of inference rules. The work to determ ine the number of inference rules demands lots of experiences from the designer. In this paper, self-organizing learning algorithms are prop osed for the fuzzy inference networks. In the proposed learning algori thms, the number of inference rules and the membership functions in th e inference rules will be automatically determined during the training procedure. The learning speed is fast. The proposed fuzzy inference n etwork (FIN) classifiers possess both the structure and the learning a bility of neural networks, and the fuzzy classification ability of fuz zy algorithms. Simulation results on fuzzy classification of two-dimen sional data are presented and compared with those of the fuzzy ARTMAP, The proposed fuzzy inference networks perform better than the fuzzy A RTMAP and need less training samples.