Table 3.

Mean (SD) data of the cross-validation results.

Model	Sensitivity	Specificity	Accuracy	Balanced accuracy	AUROCa
XGBoostb	0.8662 (0.006)	0.7655 (0.014)	0.8498 (0.004)	0.8309 (0.006)	0.9061 (0.005)
AdaBoostc	0.8546 (0.005)	0.8206 (0.012)	0.8467 (0.005)	0.8377 (0.006)	0.9076 (0.006)
GBMd	0.9110 (0.005)	0.7215 (0.001)	0.8669 (0.002)	0.8162 (0.004)	0.9114 (0.002)
LGBMe	0.8927 (0.007)	0.7462 (0.014)	0.8587 (0.002)	0.8194 (0.003)	0.083 (0.004)
RFf	0.8992 (0.006)	0.7490 (0.011)	0.8643 (0.003)	0.8241 (0.004)	0.9097 (0.002)
LRg	0.8551 (0.009)	0.8171 (0.005)	0.8259 (0.002)	0.8361 (0.003)	0.9089 (0.004)
SVMh	0.8370 (0.003)	0.8456 (0.002)	0.8390 (0.004)	0.8413 (0.009)	0.9172 (0.004)

^aAUROC: area under receiver operating characteristic.

^bXGBoost: extreme gradient boosting.

^cAdaBoost: adaptive boosting.

^dGBM: gradient boosting machine.

^eLGBM: light gradient boosting machine.

^fRF: random forest.

^gLR: logistic regression.

^hSVM: support vector machine.