Table 11 Our model’s performance in comparison to other models
From: Machine learning approaches in Covid-19 severity risk prediction in Morocco
Models | Method of reduction | Model of Prediction | Accuracy | Specifity | Sensivity (%) | AUC |
|---|---|---|---|---|---|---|
Bayat et al [56] | Features Importance | X_GBoost | 86.40% | 86.8% | 82.39 | _ |
Brinati et al [4] | _ | Random Forest | 82% | 65% | 92 | 84% |
Tschoellitsch et al [6] | Feature importance | Random Forest | 81%, | 82% | 60 | 74% |
Tordjman et al [57] | _ | Logistic Regression | _ | _ | 80.3 | 88.9% |
Soltan et al [58] | Feature importance | Extreme Gradient Boosted Tree | _ | 94.8% | 77.4 | 94% |
Alakus and Turkoglu [59] | _ | LSTM | 86.66% | _ | 99.42 | 62.50% |
Our approach | UMAP | Various Machine Learning | 100% | 100% | 100 | 100% |