Table 5.
Generalization performance comparison of ELM, RELM, and other methods presented in [3].
| Method | Testing Data | |||
|---|---|---|---|---|
| RMSE (MPa) | MAE (MPa) | MAPE (%) | R | |
| ELM | 6.0377 | 4.4419 | 15.2558 | 0.929 |
| RELM | 5.5075 | 3.9745 | 13.467 | 0.9403 |
| Individual methods [3]: | ||||
| ANN | 6.329 | 4.421 | 15.3 | 0.930 |
| CART | 9.703 | 6.815 | 24.1 | 0.840 |
| CHAID | 8.983 | 6.088 | 20.7 | 0.861 |
| LR | 11.243 | 7.867 | 29.9 | 0.779 |
| GENLIN | 11.375 | 7.867 | 29.9 | 0.779 |
| SVM | 6.911 | 4.764 | 17.3 | 0.923 |
| Ensemble methods [3]: | ||||
| ANN + CHAID | 7.028 | 4.668 | 16.2 | 0.922 |
| ANN + SVM | 6.174 | 4.236 | 15.2 | 0.939 |
| CHAID + SVM | 6.692 | 4.580 | 16.3 | 0.929 |
| ANN + SVM + CHAID | 6.231 | 4.279 | 15.2 | 0.939 |