Skip to main content
. 2026 Jan 21;12:1696945. doi: 10.3389/fmed.2025.1696945

Figure 4.

Four-panel image showing machine learning model evaluations. Panel A and B display ROC curves for train and test data with models like Logistic, Enet, MLP, DT, SVM, LightGBM, XGBoost, and KNN. Panel C shows calibration curves for models such as DT, RF, XGBoost, Enet, SVM, MLP, Logistic, LightGBM, and KNN. Panel D illustrates a decision curve analysis for test data using various models including DT, Enet, KNN, LightGBM, Logistic, MLP, RF, SVM, and XGBoost. Each graph includes different colored lines representing different algorithms.

Performance comparison of Nine machine learning models. (A) ROC curve for the training set. (B) ROC curve for the validation set. (C) Calibration curve for multiple models in the validation set. (D) Decision curve for multiple models in the validation set.