Research Article
Forecasting RMB Exchange Rate Based on a Nonlinear Combination Model of ARFIMA, SVM, and BPNN
Table 4
The RMB/USD forecasting results using ARIMA, ARFIMA, SVM, BPNN, EWL, ARIMA-SVM, ARIMA-ANN, ARFIMA-SVM, ARFIMA-ANN, NCM-r, and NCM-f.
| ā | MAE | RMSE | MAPE (%) |
| ARIMA | 3.3793 | 4.1499 | 0.5516 | ARFIMA | 2.4994 | 3.1357 | 0.4079 | SVM | 2.4324 | 2.9654 | 0.3970 | BPNN | 2.2971 | 2.9268 | 0.3751 | EWL | 2.0106 | 2.8167 | 0.3280 | ARIMA-SVM | 2.2311 | 2.8101 | 0.3641 | ARIMA-ANN | 1.9873 | 2.6342 | 0.3243 | ARFIMA-SVM | 1.8724 | 2.5280 | 0.3055 | ARFIMA-ANN | 0.8442 | 1.0254 | 0.1379 | NCM-r | 0.3910 | 0.4987 | 0.0639 | NCM-f | 0.2849 | 0.3469 | 0.0466 |
|
|