Research Article
Applicability of Artificial Neural Networks to Predict Mechanical and Permeability Properties of Volcanic Scoria-Based Concrete
Table 3
Results obtained for ANN models.
| Model | Training | Testing | Validating | R2 | RMSE | MAPE | DW | R2 | RMSE | MAPE | DW | R2 | RMSE | MAPE | DW |
| ANN1 | 0.9999 | 0.3066 | 1.2106 | 1.7605 | 0.9998 | 0.3931 | 1.5470 | 1.9505 | 0.9998 | 0.3899 | 1.7307 | 1.6328 | ANN2 | 0.9995 | 1.9783 | 2.6357 | 1.7003 | 0.9990 | 2.4439 | 4.8220 | 1.8094 | 0.9981 | 4.3322 | 4.7976 | 1.7467 | ANN3 | 0.9998 | 0.3129 | 1.4998 | 1.5035 | 0.9997 | 0.4050 | 2.0528 | 1.9018 | 0.9995 | 0.4859 | 2.6921 | 1.6489 |
|
|
ANN1: compressive strength; ANN2: water permeability; ANN3: concrete porosity.
|