Dynamic Modelling of Chaotic Time Series with Neural Networks
José Carlos Príncipe, Jyh-Ming Kuo · 1994
This paper discusses the use of artificial neural networks for dynamic modelling of time series. We argue that multistep prediction is more appropriate to capture the dynamics of the underlying dynamical system, because it constrains the iterated model. We show how this method can be implemented by a recurrent ANN trained with trajectory learning. We also show how to select the trajectory length to train the iterated predictor for the case of chaotic time series. Experimental results corroborate the proposed method. 1.0