Table 1.

Performance (CV-10 fold) of SVM-based models on both balanced and realistic datasets using AAC and DPC features as input.

Dataset	Feature	Kernel	Thr	Sen	Spec	Acc	MCC	AUC	Parameter
Balanced dataset	AAC	t0	0.1	90.14	93.66	91.9	0.84	0.94	t:0,c:80
		t1	−0.1	86.62	96.48	91.55	0.84	0.96	t:1,d:2
		t2	0.2	93.66	90.85	92.25	0.85	0.97	g:0.001:c:0.5:j:4
	DPC	t0	−0.1	90.14	89.44	89.79	0.8	0.96	t:0,c:990
		t1	−0.2	90.85	94.37	92.61	0.85	0.96	t:1,d:1
		t2	0.1	84.51	95.77	90.14	0.81	0.96	g:0.001:c:1:j:1
Realistic dataset	AAC	t0	0.3	69.72	99.3	96.62	0.78	0.95	t:0,c:5
		t1	−0.6	88.03	98.38	97.45	0.85	0.98	t:1,d:3
		t2	−0.2	86.62	98.74	97.64	0.86	0.98	g:0.001:c:10:j:1
	DPC	t0	0.5	78.87	97.89	96.17	0.77	0.95	t:0,c:990
		t1	−0.3	79.58	98.67	96.93	0.81	0.96	t:0,d:1
		t2	−0.3	83.8	98.88	97.51	0.85	0.96	g:0.001:c:1:j:2

The models constructed using balanced dataset displayed the highest MCC values of 0.85 for both AAC and DPC. The models constructed using realistic dataset displayed the highest MCC values of 0.86 and 0.85, respectively, on AAC and DPC as input features.