Fuzzy Petri nets for rule-based pattern classification

This paper proposes a new model of fuzzy Petri net for rule-based pattern classification and an algorithm to generate the network automatically. The proposed method is modified from the fuzzy min-max neural network (P. K. Simpson, IEEE Trans. on Neural Networks, vol. 3, no. 5, pp. 776-786, 1992). The modified model is modeled by the fuzzy Petri net formalism, and can be used for pattern classification. The layered model can be viewed as a collection of fuzzy production rules. This convenience makes the classification procedure transparent, as opposed to a black box as most neural network models. Both machine and human can interpret the proposed formal model for pattern classification problems. As an example of the application of the fuzzy Petri net, it is used to classify the iris data set. The result is compared with the reported model.