Modeling and Multi-Step Prediction of Chaotic Time Series Based on RBF Neural Networks

We present that RBF neural networks can be used in the modeling and prediction of chaotic time series. A three layers RBF network structure is designed and fundamental properties of RBF networks are clarified when they are used in the modeling and prediction of chaotic time series. Simulations show that RBF networks models have good fitness and high accuracy of single and multistep prediction to the chaotic time series. Using RBF networks, simulation results for modeling and prediction of chaotic time series are far better than the other methods.