Table 3.

Test set performance of each model with all features.

Model	Accuracy	AUROCa	Specificity	Sensitivity	Precision	F1 score
Random forest	0.813	0.871	0.938	0.574	0.827	0.677
Decision tree	0.705	0.674	0.772	0.577	0.568	0.572
LDAb	0.722	0.720	0.850	0.474	0.622	0.538
AdaBoostc	0.746	0.794	0.872	0.505	0.672	0.576
XGBoostd	0.674	0.763	0.913	0.213	0.559	0.309
RGFe	0.800	0.863	0.920	0.568	0.788	0.660

^aAUROC: area under the receiver operating characteristic.

^bLDA: linear discriminant analysis.

^cAdaBoost: adaptive boosting.

^dXGBoost: extreme gradient boosting.

^eRGF: regularized greedy forest.