Fig. 2From: Intelligent prediction of RBC demand in trauma patients using decision tree methodsComparison of AUC between LR, CRT and XGBoost models in predicting blood transfusion. a Non-invasive parameters to predict. b All parameters to predict. AUC area under curve, XGBoost eXtreme gradient boosting, LR logistic regression, CRT classification and regression treeBack to article page