Table 4.

The most precise prediction obtained by different intelligent estimators (1053 training and 186 testing datasets).

Model name	Datasets	MAPE%	MAE	RAPE%	RMSE	R2
LSSVR	Training data	0.25	2.86	3.94	5.64	0.99799
	Testing data	0.30	3.38	4.75	5.68	0.99794
	Training + Testing	0.26	2.94	4.06	5.65	0.99798
MLP	Training data	1.04	11.75	16.54	18.16	0.97805
	Testing data	1.10	12.47	15.68	19.98	0.97801
	Training + Testing	1.05	11.86	16.39	18.44	0.97804
CFF	Training data	1.16	13.29	18.12	18.53	0.97844
	Testing data	1.16	13.20	19.88	18.61	0.97345
	Training + Testing	1.16	13.28	18.36	18.54	0.97780
GR	Training data	0.95	10.73	14.82	16.92	0.98246
	Testing data	1.52	17.10	23.72	27.70	0.94916
	Training + Testing	1.04	11.68	16.16	18.94	0.97758
RBF	Training data	2.98	33.87	46.13	44.17	0.86954
	Testing data	2.56	29.32	44.33	38.72	0.88919
	Training + Testing	2.92	33.18	45.88	43.39	0.87158
RNN	Training data	2.52	28.57	39.49	36.93	0.90923
	Testing data	2.58	28.95	40.21	39.10	0.89494
	Training + Testing	2.53	28.63	39.59	37.26	0.90701
ANFIS	Training data	1.17	13.40	18.61	19.22	0.97605
	Testing data	1.21	13.89	18.76	20.23	0.97402
	Training + Testing	1.17	13.47	18.63	19.37	0.97573