Table 1.

Scoring metrics of the 5 ML models in predicting detonation and stability properties (D, p_C-J, Q_max, T_d, and LE) of HEDMs, provided individually for the training set and test set

Properties		Models	Training dataset			Test dataset
			RMSE	r	R2	RMSE	R	R2
Detonation performance	Qmax	XGBoost	49.694	0.984	0.958	99.988	0.914	0.825
		AdaBoost	60.579	0.970	0.938	108.252	0.898	0.794
		RF	55.731	0.983	0.948	103.574	0.926	0.812
		MLP	78.539	0.947	0.896	101.070	0.909	0.821
		KRR	96.939	0.918	0.841	101.291	0.916	0.820
	D	XGBoost	0.101	0.993	0.985	0.235	0.956	0.912
		AdaBoost	0.114	0.991	0.981	0.274	0.942	0.879
		RF	0.143	0.986	0.970	0.276	0.943	0.878
		MLP	0.185	0.976	0.949	0.244	0.959	0.905
		KRR	0.256	0.950	0.903	0.291	0.941	0.864
	pC-J	XGBoost	0.817	0.992	0.984	1.788	0.954	0.910
		AdaBoost	1.084	0.987	0.972	2.426	0.920	0.835
		RF	1.097	0.987	0.971	2.360	0.924	0.843
		MLP	0.898	0.990	0.981	2.256	0.954	0.857
		KRR	1.983	0.951	0.905	2.812	0.902	0.778
Molecular stability	Td	XGBoost	29.919	0.933	0.803	52.069	0.781	0.557
		AdaBoost	26.932	0.947	0.840	54.347	0.741	0.518
		RF	21.229	0.970	0.901	54.153	0.728	0.521
		MLP	32.256	0.878	0.770	60.670	0.635	0.399
		KRR	51.152	0.653	0.423	61.573	0.627	0.381
Crystal stability	LE	XGBoost	1.033	0.999	0.998	3.494	0.976	0.948
		AdaBoost	3.187	0.990	0.977	6.724	0.898	0.806
		RF	4.281	0.980	0.958	5.497	0.933	0.870
		MLP	3.124	0.989	0.978	6.416	0.917	0.823
		KRR	4.558	0.976	0.952	4.367	0.959	0.918

The best performing results are marked in bold.