Table 8 Accuracy for 1 year or more mortality
No | Model | Training cohort | Validation cohort | ||
|---|---|---|---|---|---|
Number of models | Accuracy (95%CI) | Number of models | ACC (95% CI) | ||
1 | LR | 12 | 0.7410 [0.7005; 0.7778] | 8 | 0.7433 [0.7128; 0.7717] |
2 | RF | 4 | 0.7982 [0.6922; 0.8744] | 8 | 0.8032 [0.7253; 0.8632] |
3 | ANN | 3 | 0.7481 [0.7252; 0.7697] | 4 | 0.7995 [0.7004; 0.8718] |
4 | DT | 4 | 0.7805 [0.6733; 0.8598] | 1 | 0.7580 [0.7521; 0.7638] |
5 | SVM | 2 | 0.6147 [0.5908; 0.6381] | 6 | 0.7105 [0.6070; 0.7959] |
6 | XGBoost | 3 | 0.8261 [0.7011; 0.9058] | 2 | 0.7431 [0.6838; 0.7946] |
7 | NB | 1 | 0.8698 [0.8499; 0.8874] | Â | Â |
8 | AdaBoost | 1 | 0.8657 [0.8603; 0.8709] | 2 | 0.8261 [0.7392; 0.8884] |
9 | KNN | 1 | 0.8304 [0.8085; 0.8502] | Â | Â |
10 | Other | Â | Â | 2 | 0.9370 [0.9208; 0.9501] |
11 | Overall | 31 | 0.7697 [0.7360; 0.8002] | 33 | 0.7837 [0.7455; 0.8175] |