Automatic optimization of pruning in evolving fuzzy neural networks using an entropy measure

In this paper we present the results of the first experiments in the investigation of automatically adjusting the learning parameters of an EFuNN. This work in part addresses previous work which speculated that this evolving connectionist system could be further developed with a view to either reducing the overall number of learning parameters or having them adjusted automatically. One of these areas is in the pruning of the EFuNN and in this case we offer an alternative method in which we apply an entropy criterion to automatically regulate the growth of it. We test this method against two benchmark classification data sets and the results of the experiments reported in this paper suggest that this new method performs better than the originally proposed method of pruning an EFuNN.