RBF network based on genetic algorithm optimization for nonlinear time series prediction

In this study, we focus on genetic algorithm (GA) for the RBF network applied to prediction of nonlinear time series. The centers and widths of the hidden layer neurons basis function - defined as the barycenter and distance between two input patterns - are coded into a chromosome. A direct inversion of matrix provides the weights between the hidden layer and output layer and avoids the risk of getting stuck into a local minimum. The performances of a network with Gaussian basis functions are compared with those of a network with genetic determination of the basis functions on the Mackey-Glass delay differential equation.