Predicting sepsis in-hospital mortality with machine learning: a multi-center study using clinical and inflammatory biomarkers

Table 2 Performances of the seven machine learning models and Sofa score for predicting in-hospital mortality

Model	AUC	Precision	Recall	F1 Score
XGBoost	0.94	0.882	0.918	0.937
Sofa score	0.687	0.849	0.879	0.914
Logistic regression	0.707	0.850	0.878	0.915
Random forest	0.686	0.852	0.882	0.917
K-nearest Neighbor	0.622	0.855	0.873	0.892
Naïve Bayes	0.590	0.842	0.876	0.914
SVM	0.648	0.855	0.873	0.892
Decision Tree	0.595	0.853	0.861	0.871

XGBoost: extreme Gradient Boosting, SVM: Support Vector Machine, AUC: the area under curve

ISSN: 2047-783X