Table 4.
Model performance of LR, MARS, CART, RF, and XGBoost methods.
| Methods | MAPE | SMAPE | RAE | RRSE | RMSE |
|---|---|---|---|---|---|
| LR | 0.1266 | 0.1226 | 0.8235 | 0.8458 | 2.3283 |
| MARS | 0.1259 | 0.1219 | 0.8228 | 0.8409 | 2.3149 |
| CART | 0.1341 | 0.1298 | 0.8666 | 0.8835 | 2.4323 |
| RF | 0.1271 | 0.1229 | 0.8222 | 0.8331 | 2.2934 |
| XGBoost | 0.1182 | 0.1155 | 0.7783 | 0.8211 | 2.2604 |
Note: LR: linear regression; MARS: multivariate adaptive regression splines; CART: classification and regression tree; RF: random forest; XGBoost: eXtreme gradient boosting; MAPE: mean absolute percentage error; SMAPE: symmetric mean absolute percentage error; RAE: relative absolute error; RRSE: root relative squared error; RMSE: root mean squared error.