局所近似法によるカオス時系列データの予測

Abstract

A Method is presented for predicting chaotic time series based on a local approximation technique. In the local approximation technique, a state space is reconstructed from a time series using delay coordinate embedding and then a local predictor is constructed on the basis of the motion of the nearest neighbors in the state space. The parameters, such as the embedding dimension and the number of the nearest neighbors,have a significant effect on the prediction accuracy. The method can be used as a means of choosing the parameters suitable for the prediction. The efficacy of the method is demonstrated using chaotic time series generated by the Hénon map and the Ikeda map.