Research Article
Coupling a Bat Algorithm with XGBoost to Estimate Reference Evapotranspiration in the Arid and Semiarid Regions of China
Table 5
Results of the statistical indicators for all models in the Yinchuan station during the training and testing stages.
| Model | Training | Testing | R2 | RMSE | MAE | RMSEmax10% | MAEmax10% | R2 | RMSE | MAE | RMSEmax10% | MAEmax10% |
| GPR1 | 0.856 | 0.653 | 0.514 | 1.371 | 1.351 | 0.87 | 0.663 | 0.511 | 1.424 | 1.401 | GPR2 | 0.982 | 0.248 | 0.191 | 0.568 | 0.548 | 0.974 | 0.334 | 0.259 | 0.73 | 0.719 | GPR3 | 0.984 | 0.23 | 0.176 | 0.53 | 0.518 | 0.974 | 0.289 | 0.197 | 0.751 | 0.696 | GPR4 | 0.98 | 0.257 | 0.196 | 0.588 | 0.574 | 0.966 | 0.343 | 0.262 | 0.752 | 0.736 | GPR5 | 0.984 | 0.23 | 0.169 | 0.536 | 0.519 | 0.972 | 0.461 | 0.378 | 0.941 | 0.907 | GPR6 | 0.986 | 0.208 | 0.147 | 0.507 | 0.491 | 0.97 | 0.384 | 0.285 | 0.888 | 0.878 | MARS1 | 0.968 | 0.3 | 0.229 | 0.677 | 0.666 | 0.974 | 0.34 | 0.262 | 0.728 | 0.723 | MARS2 | 0.988 | 0.187 | 0.142 | 0.422 | 0.409 | 0.984 | 0.262 | 0.194 | 0.609 | 0.602 | MARS3 | 0.986 | 0.203 | 0.153 | 0.472 | 0.466 | 0.984 | 0.227 | 0.184 | 0.461 | 0.453 | MARS4 | 0.992 | 0.159 | 0.12 | 0.368 | 0.36 | 0.986 | 0.219 | 0.163 | 0.523 | 0.513 | MARS5 | 0.992 | 0.155 | 0.117 | 0.348 | 0.344 | 0.99 | 0.323 | 0.266 | 0.687 | 0.676 | MARS6 | 0.996 | 0.105 | 0.07 | 0.275 | 0.265 | 0.994 | 0.172 | 0.108 | 0.464 | 0.458 | BAXGB1 | 0.978 | 0.261 | 0.189 | 0.614 | 0.594 | 0.978 | 0.261 | 0.204 | 0.58 | 0.576 | BAXGB2 | 0.988 | 0.203 | 0.146 | 0.49 | 0.468 | 0.982 | 0.229 | 0.173 | 0.524 | 0.51 | BAXGB3 | 0.99 | 0.176 | 0.13 | 0.412 | 0.402 | 0.986 | 0.19 | 0.142 | 0.416 | 0.409 | BAXGB4 | 1 | 0.051 | 0.038 | 0.116 | 0.113 | 0.988 | 0.204 | 0.156 | 0.467 | 0.452 | BAXGB5 | 0.992 | 0.205 | 0.145 | 0.502 | 0.487 | 0.988 | 0.222 | 0.158 | 0.534 | 0.532 | BAXGB6 | 0.996 | 0.162 | 0.113 | 0.393 | 0.378 | 0.988 | 0.175 | 0.123 | 0.412 | 0.401 |
|
|
Note. Bold font indicates the best model.
|