Research Article
On the Training Algorithms for Artificial Neural Network in Predicting the Shear Strength of Deep Beams
Table 5
Summary of different quality assessment criteria for the best ANN-CG model.
| | RMSE | MAE | Err. Mean | Err. Std | R | MAPE |
| Case 1: Max (R) | Training set | 12.53 | 9.82 | 0.16 | 12.61 | 0.993 | 5.79 | Testing set | 17.92 | 13.30 | 0.77 | 18.19 | 0.993 | 6.47 |
| Case 2: Min (RMSE) | Training set | 14.73 | 10.90 | −0.45 | 14.82 | 0.993 | 5.89 | Testing set | 14.02 | 11.24 | 0.12 | 14.24 | 0.992 | 6.84 |
| Case 3: Min (MAE) | Training set | 13.02 | 9.88 | 0.17 | 13.11 | 0.995 | 5.20 | Testing set | 15.49 | 10.60 | −2.47 | 15.54 | 0.981 | 6.92 |
| Case 4: Min (MAPE) | Training set | 13.48 | 10.20 | 0.10 | 13.57 | 0.992 | 6.00 | Testing set | 26.49 | 15.41 | −3.66 | 26.65 | 0.990 | 5.79 |
|
|