Application of support vector neural network with variational mode decomposition for exchange rate forecasting

摘要

A hybrid ensemble learning approach is proposed for exchange rate forecasting combining variational mode decomposition (VMD) and support vector neural network (SVNN). First, VMD is employed to decompose the original exchange rate time series into several components. Then, SVNN is adopted to forecast different component series. In the end, the forecasting results of all the components are combined using SVNN as ensemble learning method to obtain the ensemble results. Four major daily exchange rate datasets are selected for model evaluation and comparison. The empirical study demonstrates that the proposed VMD-SVNN ensemble learning approach outperforms other single forecasting models and other ensemble learning approaches in terms of both level forecasting accuracy and directional forecasting accuracy. This suggests that the VMD-SVNN ensemble learning approach is a highly promising approach for exchange rates forecasting with high volatility and irregularity.