摘要
Empirical mode decomposition (EMD) and BP_AdaBoost neural network are used in this paper to model the oil price. Based on the benefits of these two methods, we predict the oil price by using them. To a certain extent, it effectively improves the accuracy of short-term price forecasting. Forecast results of this model are compared with the results of the ARIMA model, BP neural network and EMD-BP combined model. The experimental result shows that the root mean square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE) and Theil inequality (U) of EMD and BP_AdaBoost model are lower than other models, and the combined model has better prediction accuracy.
Empirical mode decomposition (EMD) and BP_AdaBoost neural network are used in this paper to model the oil price. Based on the benefits of these two methods, we predict the oil price by using them. To a certain extent, it effectively improves the accuracy of short-term price forecasting. Forecast results of this model are compared with the results of the ARIMA model, BP neural network and EMD-BP combined model. The experimental result shows that the root mean square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE) and Theil inequality (U) of EMD and BP_AdaBoost model are lower than other models, and the combined model has better prediction accuracy.
