Table III.

Performances of the SVM classifier on the training and validation sets.

	Overall survival			ROC curve
Datasets	C-index	Brier score	Log-rank P-value	AUROC	Sensitivity	Specificity	PPV	NPV
Training set (TCGA, N=292)	0.814	0.0394	<0.0001	0.970	0.814	0.926	0.856	0.903
Validation set (GSE10141, N=62)	0.757	0.0884	0.0264	0.942	0.824	0.889	0.737	0.930

SVM, support vector machine; TCGA, The Cancer Genome Atlas; ROC, receiver operating characteristic curve; AUROC, area under receiver operating characteristic curve; PPV, positive predictive value; NPV, negative predictive value.