Prediction of bifurcation from chaotic circuit with RBF

This paper proposes a simple radial basis function (RBF) network to predict chaos and bifurcation from our original chaotic circuit. We have evaluated the prediction performance for four kinds of RBF networks and have discovered that heuristic changing of RBF network parameters can mimic the original circuit bifurcation. It suggests that RBF network is efficient not only for time series prediction but also for bifurcation prediction