Table 2.

Sensitivity and specificity averaged over 100 cross-validations.

Methods	30 features		17 features
	Sensitivity	Specificity	Sensitivity	Specificity
KNN1	0.67	0.79	0.67	0.84
KNN3	0.61	0.86	0.63	0.86
LDA	0.69	0.79	0.63	0.82
LogitBoost	0.62	0.87	0.63	0.85
NNET	0.57	0.86	0.46	0.86
PAM	0.64	0.89	0.57	0.92
QDA	NA	NA	0.18	0.98
RF	0.61	0.93	0.61	0.90
SVM.lin	0.78	0.83	0.58	0.84
SVM.rad	0.71	0.93	0.61	0.93
Tree	0.51	0.76	0.49	0.80