A Hybrid Model Based on Ensemble Empirical Mode Decomposition and Fruit Fly Optimization Algorithm for Wind Speed Forecasting

Table 3

Comparison between RBF, GRNN, and SVR and FOARBF, FOAGRNN, and FOASVR forecast for wind speed in four seasons.

	Error criteria	Spring	Summer	Fall	Winter
RBF	MAE	1.2798	0.9270	1.1633	0.9849
	RMSE	1.4989	1.1825	1.6560	1.4428
	IA	0.78923	0.6460	0.7761	0.8151
FOARBF	MAE	0.7584	0.6693	0.7583	0.7340
	RMSE	0.9144	0.8072	1.0817	1.0174
	IA	0.8653	0.8837	0.9211	0.9016
GRNN	MAE	0.8321	0.9842	1.3096	1.3101
	RMSE	1.0964	1.2857	1.5960	1.7048
	IA	0.7684	0.6164	0.6470	0.5339
FOAGRNN	MAE	0.7371	0.6912	0.7296	0.7186
	RMSE	0.8881	0.8404	1.0394	0.9933
	IA	0.8738	0.8669	0.9245	0.9016
SVR	MAE	1.0776	1.0346	1.3319	2.6280
	RMSE	1.2551	1.3142	1.8932	4.2264
	IA	0.8033	0.7448	0.7526	0.5128
FOASVR	MAE	0.7440	0.6319	0.6941	0.6798
	RMSE	0.8755	0.7812	0.9697	0.9799
	IA	0.8740	0.8914	0.9346	0.9097