Table 4.

Performance comparison of different machine learning models.

Models	RMSE (%)		R2		MAE
	Calibration Sets	Test Sets	Calibration Sets	Test Sets	Calibration Sets	Test Sets
PLS	14.36	13.34	0.80	0.81	11.18	10.69
Ridge	17.09	14.84	0.74	0.78	13.39	11.81
Enet	15.23	14.38	0.77	0.78	12.11	11.60
Rqlasso	15.72	14.92	0.76	0.77	12.46	11.94
Earth	16.30	16.84	0.74	0.71	12.93	13.14
Kknn	16.44	16.02	0.75	0.74	12.79	12.38
ParRF	15.91	14.87	0.77	0.79	12.92	11.91
Qrf	15.66	14.81	0.76	0.77	11.99	10.99
Rf	15.92	14.99	0.77	0.78	12.95	11.98
Ctree	21.74	22.71	0.55	0.48	17.09	16.95
Cubist	12.67	10.93	0.84	0.87	9.78	8.37
Glmboost	15.20	14.38	0.77	0.78	12.17	11.57
XgbTree	29.67	29.22	0.33	0.30	22.70	22.86
Msaene	15.33	14.39	0.77	0.78	12.37	11.69

MAE was the average absolute error.