Skip to main content
. 2026 Jun 4;13:1808657. doi: 10.3389/fmed.2026.1808657

Figure 3.

Panel a shows an ROC curve comparing six machine learning models for classification, panel b presents a calibration plot with predicted versus actual probabilities, and panel c displays a decision curve showing net benefit across prediction thresholds, with each model represented by a distinct color-coded line.

Comparison of the performance of six algorithms on the test set. (a) The first figure represents the ROC curve. (b) The second figure represents the calibration curve. (c) The third figure represents the DCA decision curve. AUC = area under curve, KNN = K-nearest neighbors, SVM = support vector machine, LR = logistic regression, VTE = venous thromboembolism.