Table 5.
The results of the proposed methods used to classify COVID-19 patients.
| Random Forests | XGBoost | Ada Boost | AI-Score | |
|---|---|---|---|---|
| Acc_severe (%) | 96.84 | 98.94 | 91.57 | 98.94 |
| Se_severe (%) | 91.42 | 97.14 | 77.14 | 97.14 |
| Sp_severe (%) | 100 | 100 | 100 | 100 |
| PPV_severe (%) | 100 | 100 | 100 | 100 |
| NPV_severe (%) | 95.23 | 98.36 | 88.23 | 98.36 |
| Acc_moderate (%) | 97.89 | 97.89 | 90.52 | 98.94 |
| Se_moderate (%) | 93.33 | 93.33 | 100 | 96.66 |
| Sp_moderate (%) | 100 | 100 | 86.15 | 100 |
| PPV_moderate (%) | 100 | 100 | 76.92 | 100 |
| NPV_moderate (%) | 97.01 | 97.01 | 100 | 98.48 |
| Acc_mild (%) | 94.73 | 96.84 | 98.94 | 97.89 |
| Se_mild (%) | 100 | 100 | 96.66 | 100 |
| Sp_mild (%) | 92.30 | 95.38 | 100 | 96.92 |
| PPV_mild (%) | 85.71 | 90.90 | 100 | 93.75 |
| NPV_mild (%) | 100 | 100 | 98.48 | 100 |
Acc: Accuracy; Se: Sensitivity; Sp: Specificity; PPV: Positive Predictive Value; NPV: Negative Predictive Value.