Figure 2. Machine Learning Model Performance.

Receiver operating characteristic curves and precision recall curves for the deep learning and baseline models (logistic regression, support vector machine [SVM], random forest, and XGBoost) in data sets I and II. A, Receiver operating characteristic curve (left panel) and precision recall curve (right panel) for data set I (4950 sections). B, Receiver operating characteristic curve (left panel) and precision recall curve (right panel) for data set II (2000 sections). In the left panels, gray-dashed curves contain all points in the precision/recall space whose F1 scores are the same, and the gray-dotted diagonal line indicates the performance based on random guessing.