Price forecasting of stock index futures based on a new hybrid EMD-RBF neural network model

The price of stock index futures in China has exhibited non-stationary and non-linear signal characteristics in the last four years. Traditional prediction methods cannot accurately predict the long correlation sequence. Therefore, a new forecasting method for the settlement price forecast of stock index futures in China is established by simultaneously using empirical mode decomposition and radial basis function. Results show that the original sequence with long correlation properties is decomposed into several frequency bands with short correlation properties, thereby solving the shortcoming of the original sequence being stochastic. System dynamic information is insufficient because of the interference of adjacent frequency bands. Compared with other forecasting models, the proposed hybrid model exhibits higher prediction accuracy.