A hybrid MGA-BP algorithm for RBFNs self-generate

This paper proposes a novel hybrid algorithm to determine the parameters (number of neurons, centers, widths and weights) of radial basis function neural networks automatically. In this work, a hybrid algorithm combines the multi-encoding genetic algorithm (MGA) and the back propagation (BP) algorithm to form a hybrid learning algorithm (MGA-BP) for training radial basis function networks (RBFNs), which adapts to the network structure and updates its weights by choosing a special fitness function. The proposed method is used to deal with non-linear identification problems, and the results obtained are compared with existent bibliography, showing an improvement over the published methods.