Forecasting RMB Exchange Rate Based on a Nonlinear Combination Model of ARFIMA, SVM, and BPNN

Table 5

The RMB/EUR forecasting results using ARIMA, ARFIMA, SVM, BPNN, EWL, ARIMA-SVM, ARIMA-ANN, ARFIMA-SVM, ARFIMA-ANN, NCM-r, and NCM-f.

	MAE	RMSE	MAPE (%)
ARIMA	9.7196	12.6395	1.1497
ARFIMA	9.1906	11.9829	1.0872
SVM	7.4141	9.6193	0.8772
BPNN	6.1349	7.8148	0.7263
EWL	7.1599	9.2979	0.8473
ARIMA-SVM	6.0254	7.3706	0.7171
ARIMA-ANN	5.5129	7.0651	0.6528
ARFIMA-SVM	5.4177	6.2094	0.6458
ARFIMA-ANN	4.8081	5.9508	0.5719
NCM-r	2.6946	3.2667	0.3210
NCM-f	2.6097	3.1159	0.3111