Table 3.
Experimental results of the state-of-the-art algorithms for prediction of confirmed cases on global COVID-19 datasets.
| Model | MAE | MSE | RMSE | R2 | RMSLE | MAPE |
|---|---|---|---|---|---|---|
| K Neighbors Regressor | 369.837 | 1.49381e+07 | 3782.07 | 0.0456 | 1.501 | 4.1186 |
| Extra Trees Regressor | 365.563 | 1.51515e+07 | 3811.82 | 0.0307 | 1.3096 | 3.3704 |
| Random Forest | 368.821 | 1.52086e+07 | 3823.22 | 0.0245 | 1.3027 | 3.3131 |
| Decision Tree | 385.084 | 1.52274e+07 | 3819.96 | 0.0162 | 1.4723 | 7.605 |
| Support Vector Machine | 374.921 | 1.54591e+07 | 3853.44 | 0.01 | 1.5798 | 4.0062 |
| Huber Regressor | 380.769 | 1.56759e+07 | 3882.29 | -0.0054 | 1.8197 | 2.9956 |
| Ridge Regression | 383.174 | 1.56992e+07 | 3885.28 | -0.007 | 1.7949 | 2.0534 |
| Least Angle Regression | 383.169 | 1.56992e+07 | 3885.28 | -0.007 | 1.7949 | 2.0548 |
| Linear Regression | 383.169 | 1.56992e+07 | 3885.28 | -0.007 | 1.7949 | 2.0548 |
| Bayesian Ridge | 383.242 | 1.56999e+07 | 3885.36 | -0.0071 | 1.7958 | 2.0343 |
| AdaBoost Regressor | 385.502 | 1.5716e+07 | 3887.52 | -0.0083 | 1.7628 | 1.4017 |
| Orthogonal Matching Pursuit | 386.552 | 1.5721e+07 | 3888.17 | -0.0086 | 1.8743 | 1.6181 |
| Lasso Regression | 391.905 | 1.57419e+07 | 3890.79 | -0.01 | 2.4943 | 0.8246 |
| Elastic Net | 391.69 | 1.57415e+07 | 3890.73 | -0.01 | 2.4081 | 0.8149 |
| Lasso Least Angle Regression | 391.905 | 1.57419e+07 | 3890.79 | -0.01 | 2.4943 | 0.8246 |
| CatBoost Regressor | 482.418 | 9.60296e+07 | 6272.84 | -3.5039 | 1.3871 | 2.6725 |
| Light Gradient Boosting Machine | 474.62 | 7.08946e+07 | 6155.48 | -7.7306 | 1.3274 | 2.6168 |
| Extreme Gradient Boosting | 5618.07 | 1.96674e+11 | 143,720 | -13574.3 | 1.5256 | 2.5724 |
| Passive Aggressive Regressor | 7795.09 | 3.02021e+11 | 184,794 | -20857.5 | 2.5097 | 95.2851 |
| Gradient Boosting Regressor | 8468.15 | 3.52228e+11 | 191,742 | -35165.9 | 1.5391 | 4.9954 |