Forecasting Volatility of Chinese Composite Index Based on Empirical Mode Decomposition and Neural Network

Empirical Mode Decomposition (EMD), recently proposed by Huang et al. [12], appears to be a novel data analysis method for nonlinear and non-stationary time series. By decomposing a time series into a small number of independent and concretely implicational intrinsic modes based on scale separation, EMD explains the generation of time series data from a novel perspective. This paper presents an empirical mode decomposition based on neural network learning paradigm (EMD-NN) for forecasting volatilities of Shanghai A shares (Shanghai) and Shenzhen A shares (Shenzhen). By the criteria of some statistic loss functions, EMD-NN outperforms GARCH family models (GARCH, EGARCH, GJR), moving average and neural network in improving predictive accuracy. Keywords-Empirical Mode Decomposition; GARCH; EGARCH; GJR; Moving average; Neural network.