Skip to main content
. 2020 Nov 20;10:20299. doi: 10.1038/s41598-020-75352-7

Figure 3.

Figure 3

Performances of models trained by various methods (random prediction, logistic regression, SVM, RF, XGBoost, and ANN): (a) by accuracy, (b) by the AUC of the ROC curve from the sensitivity–specificity plot, (c) by F1 score, (d) by the AUC of the precision-recall curve.