Table 10 Performance of the developed models in comparison with the compositional models.

Models	Train			Test			Overall
Models	RMSE	R²	AARD (%)	RMSE	R²	AARD (%)	RMSE	R²	AARD (%)
XGBoost^a	0.0046	0.9997	0.2085	0.0111	0.9980	0.4646	0.0059	0.9994	0.2598
GradientBoosting^a	0.0057	0.9996	0.2948	0.0159	0.9960	0.6110	0.0078	0.9989	0.3581
CatBoost^a	0.0105	0.9987	0.5293	0.0154	0.9962	0.9603	0.0114	0.9982	0.5615
Normal Random Forest^b	0.0425	0.9866	0.9390	0.0541	0.9745	1.0424	0.0451	0.9844	0.9597
Normal Decision Trees^b	0.0645	0.9703	1.2312	0.0430	0.9797	1.4002	0.0608	0.9717	1.2650
Normal Extra Trees^b	0.0261	0.9944	1.2132	0.0342	0.9929	1.3511	0.0279	0.9940	1.2408
Lumped Random Forest^b	0.0395	0.9894	0.9600	0.0250	0.9898	1.0426	0.0370	0.9895	0.9766
Lumped Decision Trees^b	0.0966	0.9293	1.3422	0.0510	0.9793	1.4343	0.0893	0.9389	1.3607
Lumped Extra Trees^b	0.0248	0.9954	1.1404	0.0320	0.9915	1.2785	0.0264	0.9947	1.1681

Quick links

Search