Application of Independent Component Analysis Preprocessing and Support Vector Regression in Time Series Prediction

In this study, the application of independent component analysis (ICA), a new feature extraction method, and support vector regression (SVR) in time series prediction is presented. The proposed method first use ICA as preprocessing to transform the input space composed of original time series data into the feature space consisting of independent components (ICs) representing underlying information/features of the original data. Then, prediction models will be built by using SVR for ICs. Finally, the predicted value of each IC will be transformed back into the original space for time series prediction. Experimental results on the forecasting of NTD/USD exchange rate have showed that the proposed method outperforms the SVR model without ICA preprocessing.