Multiperiod-Ahead Wind Speed Forecasting Using Deep Neural Architecture and Ensemble Learning

Table 3

Performance criteria of forecasting models.

	The proposed model					The single SDAE-ELM
Indices	RMSE	MAE	BIAS	SDE	APE	RMSE	MAE	BIAS	SDE	APE
15-min	1.30	1.11	-1.00	0.83	0.64	2.19	1.79	-1.26	1.80	1.23
1-h	1.38	0.98	-0.67	0.76	0.68	1.59	1.26	-1.12	1.21	0.87
4-h	1.54	1.33	-0.87	1.27	0.96	1.67	1.68	-1.59	1.40	1.35
8-h	1.98	1.52	-0.9	1.99	1.24	2.22	1.95	-1.30	2.12	1.29
24-h	1.87	1.42	-0.09	1.87	0.30	2.64	2.24	-1.30	2.30	0.51
	ELMAN					ANFIS
Indices	RMSE	MAE	BIAS	SDE	APE	RMSE	MAE	BIAS	SDE	APE
15-min	4.28	3.29	0.10	4.28	3.43	4.13	3.13	0.22	4.12	3.29
1-h	4.52	3.35	0.98	4.76	3.58	4.38	2.86	0.88	3.74	2.98
4-h	4.67	3.59	1.76	4.69	3.68	4.88	4.09	1.34	4.58	3.76
8-h	5.12	3.97	-1.65	5.12	3.98	5.34	4.28	-1.59	5.13	4.01
24-h	1.91	1.56	-1.08	1.58	0.44	2.29	1.82	0.92	2.09	0.50