Study of prediction based on RBF Neural network optimized by PSO

Abstract

The parameters of RBF neural network have important effect to its performance. The parameters selection is the important research content of the RBF Neural network. For this problem, this study proposed a kind of method to choose the parameters of the RBF neural network by particle swarm optimization algorithm (PSO). The experiment result indicates the RBF neural network prediction model optimized by PSO has high prediction accuracy, and PSO is one kind of effective method for RBF neural network parameters selection.