Table 3.

Performance of each classifier based on the candidate feature set in the primary and validation cohorts

Classifier	AUC	Sensitivity	Specificity	Accuracy
Primary cohort
LR	0.916 (0.885–0.938)	67.6% (25/37)	90.4% (350/387)	0.884 (0.851–0.911)
SVM-Linear	0.803 (0.760–0.838)	51.4% (19/37)	86.0% (333/387)	0.830 (0.790–0.864)
SVM-RBF	0.821 (0.780–0.856)	75.7% (28/37)	84.0% (325/387)	0.833 (0.793–0.866)
RF	0.924 (0.894–0.947)	59.5% (22/37)	93.0% (360/387)	0.901 (0.867–0.927)
XGBoost	0.964 (0.941–0.979)	75.7% (28/37)	96.4% (373/387)	0.946 (0.919–0.965)
Validation cohort
XGBoost	0.974 (0.910–0.996)	100% (8/8)	85.6% (77/90)	0.867 (0.780–0.925)

AUC, area under the receiver operating characteristic curve; LR, logistic regression; RF, random forest; SVM-Linear, support vector machine with a linear kernel; SVM-RBF, support vector machine with a radial basis function; XGBoost, extreme gradient boosting