Optimal Polynomial Fuzzy Swarm Net for Handling Data Classification Problems

In this paper, we introduce a new topology of optimal polynomial fuzzy swarm net (OPFSN) that is based on swarm optimized multilayer perceptron with fuzzy polynomial neurons. The study offers a comprehensive design methodology involving mechanisms of particle swarm optimization (PSO). The design of the conventional PNN uses extended group methods of data handling (GMDH) with a fixed scheme for the network. It also considers a fixed number of input nodes in each layer and the resulting architecture does not guarantee optimal network architecture. Here, the development of OPFSN gives rise to a structurally optimized topology and comes with a substantial level of flexibility which becomes apparent when contrasted with the one we encounter in the conventional PNN. To evaluate the performance of the swarm optimized OPFSN, we experimented with bench mark data sets. A comparative analysis reveals that the proposed OPFSN exhibits higher classification accuracy in comparison to PNN.