Table 2. Comparison of the predictive performance of the five models in terms of mean errors (ME), mean absolute errors (MAE), and mean relative errors (MRE).

The best indicator values are in bold.

	ME	MAE	MRE
Model 1: LM	−0.00186	0.00401	1.26
Model 2: FEM	−0.00105	0.00507	1.60
Model 3: LMCCC	−0.00094	0.00476	1.50
Model 4: rLMCCC	0.00047	0.00600	1.89
Model 5: LMVCC	−0.00176	0.00363	1.14