Table 5 Accuracy for in-hospital mortality
No | Model | Training cohort | Validation cohort | ||
|---|---|---|---|---|---|
Number of models | Accuracy (95% CI) | Number of models | ACC (95% CI) | ||
1 | LR | 15 | 0.8302 [0.7800; 0.8709] | 16 | 0.8487 [0.8250; 0.8697] |
2 | RF | 6 | 0.8106 [0.7462; 0.8617] | 12 | 0.8881 [0.8639; 0.9084] |
3 | ANN | 3 | 0.8864 [0.7927; 0.9409] | 2 | 0.8651 [0.8364; 0.8893] |
4 | DT | 4 | 0.8824 [0.8094; 0.9299] | 2 | 0.8464 [0.7164; 0.9232] |
5 | SVM | 4 | 0.7902 [0.7333; 0.8376] | 6 | 0.8199 [0.7772; 0.8559] |
6 | XGBoost | 6 | 0.8679 [0.8033; 0.9136] | 4 | 0.8505 [0.8332; 0.8663] |
7 | NB | 1 | 0.7984 [0.7892; 0.8073] | 2 | 0.7615 [0.7342; 0.7869] |
8 | KNN | Â | Â | 1 | 0.9112 [0.8973; 0.9233] |
9 | Other | 3 | 0.8800 [0.7551; 0.9458] | Â | Â |
10 | Overall | 42 | 0.8434 [0.8166; 0.8669] | 45 | 0.8569 [0.8411; 0.8715] |