Table 2.

RMSE values for different forecast time horizons.

Data Set	ANN Model	tHP = 1h	tHP = 2h	tHP = 3h	tHP = 4h
Learning	ann01	2.10	2.54	2.93	3.41
	ann02	2.01	2.54	2.81	3.38
	ann03	2.63	2.95	3.38	4.01
	ann04	2.15	2.48	2.74	3.23
	ann05	2.37	2.79	3.22	4.01
	ann06	2.75	3.01	3.36	4.10
	ann07	2.19	2.68	3.00	3.59
	ann08	2.34	2.91	3.34	3.80
	ann09	2.21	2.93	3.40	4.16
	ann10	2.16	2.67	3.02	3.39
Validation	ann01	2.36	2.87	3.17	3.63
	ann02	2.28	2.97	3.30	3.49
	ann03	2.47	2.83	3.13	3.60
	ann04	2.37	2.78	3.17	3.66
	ann05	2.40	2.78	3.10	3.63
	ann06	2.50	2.83	3.08	3.53
	ann07	2.49	2.90	3.12	3.39
	ann08	2.45	2.74	3.15	3.50
	ann09	2.19	2.81	3.13	3.51
	ann10	2.36	2.68	3.08	3.50
Testing	ann01	2.57	3.41	4.12	4.65
	ann02	2.53	3.39	4.07	4.59
	ann03	2.80	3.38	3.94	4.48
	ann04	2.53	3.35	4.22	5.01
	ann05	2.61	3.34	4.10	4.82
	ann06	2.96	3.61	4.28	4.96
	ann07	2.83	3.42	4.09	4.70
	ann08	2.90	3.37	4.15	4.95
	ann09	2.52	3.40	4.11	4.81
	ann10	2.81	3.29	4.07	4.87

RMSE: root mean-square error; t_HP: time horizon of prediction.