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

Table 4

The forecasting results of model selection among the FOARBF, FOAGRNN, and FOASVR in spring.

Components	Error criteria	FOARBF	FOAGRNN	FOASVR
IMF2	MAE	0.1679	0.1330	0.0769
	RMSE	0.1935	0.1653	0.0945
	IA	0.9013	0.9307	0.9808
IMF3	MAE	0.0879	0.0762	0.0452
	RMSE	0.1089	0.0947	0.0599
	IA	0.9872	0.9900	0.9963
IMF4	MAE	0.1297	0.0603	0.0766
	RMSE	0.1604	0.0717	0.0878
	IA	0.9321	0.9867	0.9751
IMF5	MAE	0.0422	0.1298	0.1514
	RMSE	0.0595	0.1602	0.1727
	IA	0.9992	0.9949	0.9932
IMF6	MAE	0.4546	0.2836	0.0052
	RMSE	0.6196	0.3994	0.0103
	IA	0.7801	0.9034	1.0000
IMF7	MAE	0.0429	0.1394	0.1276
	RMSE	0.0433	0.1399	0.1354
	IA	0.9976	0.9754	0.9794
	MAE	0.2081	0.0025	0.0178
	RMSE	0.2081	0.0026	0.0304
	IA	0.4322	0.9998	0.9614