Table 3. Predictive performances of the 3 machine learning models.

Model	MSEa	RMSEb	R²c	MAEd	MAPEe
RFf	11.01	3.32	0.71	2.49	0.48
XGBoostg	14.98	3.87	0.61	2.87	0.53
DTh	26.88	5.18	0.29	3.92	0.59

^aMSE: mean square error.

^bRMSE: root-mean-squared error.

^cR²: coefficient of determination.

^dMAE: mean absolute error.

^eMAPE: mean absolute percentage error.

^fRF: random forest.

^gXGBoost: extreme gradient boosting.

^hDT: decision tree.