Table 3.
Performance of the machine learning approaches for the estimation of use of positive inotropic agents.
| Model | AUCa | Accuracy | Precision | Recall | F1 | Brier score |
| LRb | 0.85 | 0.78 | 0.77 | 0.79 | 0.78 | 0.16 |
| RFc | 0.96 | 0.87 | 0.85 | 0.91 | 0.88 | 0.10 |
| SVMd | 0.91 | 0.85 | 0.83 | 0.88 | 0.84 | 0.17 |
| ANNe | 0.90 | 0.83 | 0.78 | 0.94 | 0.84 | 0.12 |
| XGBoostf | 0.96 | 0.84 | 0.79 | 0.94 | 0.86 | 0.11 |
aAUC: area under the curve.
bLR: logistic regression.
cRF: random forest.
dSVM: support vector machine.
eANN: artificial neural network.
fXGBoost: extreme gradient boosting.