Skip to main content
Fig. 4 | Radiation Oncology

Fig. 4

From: Machine learning-based normal tissue complication probability model for predicting albumin-bilirubin (ALBI) grade increase in hepatocellular carcinoma patients

Fig. 4

External validation (N = 44) performances of the best gradient-boosted tree model and ensemble model. (a) ROC curves for the simple model using only the baseline ALBI score as input (ALBI only, gray dashed line), the best gradient-boosted tree model (GBT, solid blue line), and the ensemble of logistic regression and gradient-boosted tree model (Ensemble, solid orange line). The AUROC is indicated for each model. (b) Precision-recall curves for the models shown in A). Average precision (AP) was indicated for each model

Back to article page