Combining Artificial Intelligence with Non-linear Data Processing Techniques for Forecasting Exchange Rate Time Series

Abstract

Combing back-propagation neural network (BPNN) and empirical mode decomposition (EMD) techniques, this study proposes EMD-BPNN model for forecasting. In the first stage, the original exchange rate series were first decomposed into a finite, and often small, number of intrinsic mode functions (IMFs). In the second stage, kernel predictors such as BPNN were constructed for forecasting. Compared with traditional model (random walk), the proposed model performs best. This study significantly reduced errors not only in the derivation performance, but also in the direction performance.