A Novel Hybrid Intelligent Model for Financial Time Series Forecasting and Its Application

Due to the fluctuation and complexity of the financial time series, it is difficult to use any single artificial technique to capture its non-stationary property and accurately describe its moving tendency. So a novel hybrid intelligent forecasting model based on empirical mode decomposition (EMD) and support vector regression (SVR) is proposed. EMD can adaptively decompose the complicated raw data into a finite set of intrinsic mode functions (IMFs) and a residue, which have simpler frequency components and higher correlation. Tendencies of these IMFs and the residue are forecasted by SVR respectively, in which the kernel functions are appropriately chosen according to their different fluctuations. The final forecasting value can be obtained by the sum of these prediction results. Successful forecasting application of Shanghai-securities index demonstrates the feasibility and validity of the presented model.