Prediction of Chaotic Time Series Using LS-SVM with Automatic Parameter Selection

Least squares support vector machine (LS-SVM) combined with genetic algorithm (GA) is used to predict chaotic time series. The LS-SVM can overcome some shortcoming in the multilayer perceptron and the GA is used to tune the LS-SVM parameters automatically. A benchmark problem, Hénon map time series, has been used as an example for demonstration. It is showed this approach can escape from the blindness of man-made choice of the LS-SVM parameters. It enhances the efficiency and the capability of prediction. Further, the GA is compared with cross-validation method for tuning LS-SVM parameters. The results reveal that the GA can obtain lower prediction errors than the k-folds cross validation method.