Table 5.

Model performance without the clinical assessment features.

Model name	Accuracy	Precision	Recall	F1-score	Specificity	AUROCa
LRb	0.720	0.350	0.870	0.610	0.690	0.814
SVMc	0.840	0.600	0.070	0.335	0.990	0.824
KNNd	0.800	0.430	0.700	0.565	0.820	0.821
Decision tree	0.835	0.494	0.530	0.512	0.894	0.712
Random forest	0.868	0.610	0.501	0.556	0.938	0.893
XGBooste	0.890	0.720	0.500	0.610	0.960	0.868
AdaBoostf	0.830	0.470	0.570	0.520	0.880	0.815
DNNg	0.893	0.687	0.632	0.660	0.944	0.899

^aAUROC: area under the receiver operating characteristic curve.

^bLR: logistic regression.

^cSVM: support vector machine.

^dKNN: k-nearest neighbor.

^eXGBoost: extreme gradient boost.

^fAdaBoost: adaptive boosting.

^gDNN: deep neural network.