A fuzzy neural network for system modeling

In this paper, an incrementally generated fuzzy neural network (FNN) for fuzzy system modeling is described. This FNN combines the features of initial fuzzy model self-generation, fast input selection, partition validation, parameter optimisation and rule-base simplification. Experimental studies demonstrate that the FNN is able to achieve accuracy comparable to or higher than both a feedforward crisp neural network i.e. NeuroRule, and a decision tree i.e. C4.5, with more compact rule bases for most of the data sets used in our experiment. In addition, the FNN is insensitive to the problem of small disjuncts that affects decision trees. This can be very useful in real-world situations, since the data of interest can often be only a small fraction of the available data.