Skip to main content

View full-text article in PMC

. 2020 Jul 6;8(7):e17257. doi: 10.2196/17257

Table 4.

Prediction scores of models created by different algorithms.

Algorithm/ model	AUC^a	ACC^b	F1-score	Sensitivity	PPV^c	Specificity	NPV^d
Logistic regression	0.865	0.809	0.785	0.736	0.840	0.874	0.787
Decision tree	0.882	0.827	0.802	0.742	0.873	0.903	0.796
KNN^e	0.908	0.827	0.808	0.769	0.851	0.879	0.810
SVM^f	0.915	0.850	0.832	0.782	0.888	0.912	0.824
Random forest	0.938	0.861	0.846	0.812	0.884	0.905	0.843
XGBoost	0.943	0.870	0.855	0.820	0.895	0.914	0.849

^aAUC: area under the receiver operating curve.

^bACC: accuracy.

^cPPV: positive predictive value.

^dNPV: negative predictive value.

^eKNN: K-nearest neighbor.

^fSVM: support vector machine.