Skip to main content
. 2021 Mar 10;6(6):1–9. doi: 10.1001/jamacardio.2021.0122

Figure 2. Predicted Risk of In-Hospital Mortality by Machine Learning and Logistic Regression Models.

Figure 2.

Extreme gradient boosting model (XGBoost) (A), neural network (B), and meta-classifier model (C), using the 29-variable input used in the development of the model by McNamara et al.21 The shaded areas denote standard error of the calibration.