Research Article
Applying Adaptive Neural Fuzzy Inference System to Improve Concrete Strength Estimation in Ultrasonic Pulse Velocity Tests
Table 4
Adaptive neuro-fuzzy inference system (ANFIS) model.
| Attribution function | MAPE (%) | MAPE (%) | One-input: number of functions | Two-input: number of functions | [2] | [3] | [4] | [5] | [2 2] | [3 3] | [4 4] | [5 5] |
| Trimf | 10.04 | 9.80 | 9.66 | 9.80 | 9.98 | 9.94 | 10.05 | 10.46 | Trapmf | 10.26 | 10.94 | 9.48 | 10.00 | 10.52 | 10.94 | 10.43 | 9.18 | Gbellmf | 9.73 | 9.89 | 9.46 | 9.37 | 9.68 | 9.89 | 10.40 | 9.56 | Gaussmf | 9.76 | 10.01 | 9.41 | 9.76 | 9.77 | 10.85 | 10.28 | 9.94 | Gauss2mf | 9.78 | 10.01 | 9.49 | 9.88 | 10.18 | 10.54 | 10.26 | 10.98 | Pimf | 10.48 | 10.50 | 10.50 | 10.50 | 10.73 | 10.66 | 10.31 | 10.49 | Dsigmf | 10.15 | 9.97 | 9.46 | 9.69 | 10.11 | 10.56 | 10.67 | 10.58 | Psigmf | 10.15 | 9.91 | 9.46 | 9.43 | 10.11 | 10.57 | 10.67 | 10.56 | Trimf | 10.04 | 9.80 | 9.66 | 9.80 | 9.98 | 9.94 | 10.05 | 10.46 | Trapmf | 10.26 | 10.94 | 9.48 | 10.00 | 10.52 | 10.94 | 10.43 | 9.18 |
|
|