Table 1.

The classification accuracy results using training set with various combinations of ${\nu }_{s1}$ and ${\nu }_{s2}$ (degrees of freedom of $t$ in the first and second stage algorithms, respectively) and $\beta$ (the smoothness parameter) with $k_{\mathrm{ext}}=20$ and $k_{\mathrm{int}}=1$.

${\nu }_{s1}$	${\nu }_{s2}$	$\beta$	Accuracy
			Sensitivity (%)	Specificity (%)	AUC
4	4	5	88	64	0.90
4	4	10	85	82	0.90
4	4	15	85	94	0.91
4	4	20	77	97	0.88
4	4	25	77	97	0.88
4	4	30	77	97	0.88
4	10	5	88	61	0.89
4	10	10	88	67	0.90
4	10	15	85	82	0.90
4	10	20	85	88	0.90
4	10	25	81	91	0.89
4	10	30	77	91	0.87
4	20	5	88	61	0.90
4	20	10	88	64	0.89
4	20	15	88	70	0.90
4	20	20	85	70	0.89
4	20	25	85	82	0.90
4	20	30	85	88	0.91
10	10	5	85	42	0.81
10	10	10	85	55	0.83
10	10	15	81	70	0.84
10	10	20	77	76	0.83
10	10	25	73	79	0.82
10	10	30	69	79	0.80
10	20	5	88	42	0.82
10	20	10	85	52	0.83
10	20	15	85	61	0.84
10	20	20	81	61	0.83
10	20	25	77	67	0.82
10	20	30	77	73	0.82
20	20	5	92	36	0.82
20	20	10	88	42	0.82
20	20	15	88	48	0.83
20	20	20	81	55	0.80
20	20	25	81	61	0.81
20	20	30	81	67	0.82