Table 3.

Performance of optimized single-encodings and fusion models on different machine learning classifiers.

GBDT SVM NB AB RF
Descriptor	Acc	Sp	Sn	MCC	AUC	Acc	Sp	Sn	MCC	AUC	Acc	Sp	Sn	MCC	AUC	Acc	Sp	Sn	MCC	AUC	Acc	Sp	Sn	MCC	AUC
AAC	76.34	74.70	79.60	0.520	0.827	55.39	55.20	57.50	0.108	0.551	74.55	77.00	70.10	0.493	0.800	76.35	74.70	79.60	0.528	0.811	76.05	74.65	77.68	0.518	0.824
CKSAAP	51.19	52.00	79.60	0.029	0.526	50.29	50.20	60.50	0.006	0.503	52.09	54.00	24.20	0.051	0.523	50.09	50.00	80.20	0.001	0.505	53.22	53.12	79.88	0.032	0.551
C/T/D	82.33	81.40	83.80	0.647	0.896	58.68	55.00	91.00	0.231	0.587	72.45	68.00	83.80	0.461	0.808	75.44	73.50	79.50	0.511	0.820	82.16	81.10	83.24	0.644	0.892
Geary	78.74	77.30	81.40	0.576	0.854	50.39	52.70	47.30	0.048	0.524	74.55	75.90	71.90	0.492	0.816	69.76	70.10	68.90	0.395	0.762	79.71	78.10	81.77	0.582	0.876
NMBroto	81.13	80.60	82.00	0.623	0.890	79.04	80.50	76.60	0.582	0.790	79.94	81.60	77.20	0.600	0.877	68.26	67.00	71.90	0.366	0.765	80.93	79.83	82.00	0.621	0.890
PAAC	76.04	74.90	78.40	0.522	0.825	70.65	67.50	79.60	0.420	0.707	62.87	58.60	88.00	0.298	0.745	69.16	67.00	75.40	0.386	0.754	75.88	74.40	78.32	0.521	0.825
TPC	50.89	80.00	24.00	0.074	0.506	49.40	49.50	59.30	−0.01	0.492	50.59	51.40	21.60	0.015	0.502	49.40	49.30	39.50	−0.01	0.493	50.28	78.90	24.00	0.071	0.498
Fusion All	93.06	94.00	92.00	0.862	0.971	79.34	78.40	80.91	0.574	0.794	87.00	88.90	84.70	0.752	0.936	82.33	76.30	87.50	0.628	0.888	91.08	92.55	91.11	0.837	0.956