Research Article
Is the Linear Modeling Technique Good Enough for Optimal Form Design? A Comparison of Quantitative Analysis Models
Table 7
Predicted image values and RMSE of NN models for the test set.
| | PDA no. | 25 | 26 | 27 | 28 | 29 | 30 | RMSE |
| | NN-FE-P | 5.08 | 6.28 | 3.57 | 4.18 | 5.32 | 8.22 | 0.2663 | 0.3033 | | NN-FE-S | 4.43 | 5.04 | 3.53 | 5.20 | 9.23 | 8.76 | 0.3565 | | NN-FE-C | 1.45 | 2.95 | 3.71 | 9.03 | 9.38 | 6.65 | 0.3701 | | NN-FE-N | 4.63 | 6.04 | 3.26 | 3.67 | 5.25 | 5.41 | 0.2203 |
| | NN-FT-P | 2.50 | 8.27 | 3.44 | 6.04 | 3.97 | 2.69 | 0.2875 | 0.3168 | | NN-FT-S | 1.58 | 8.27 | 3.41 | 1.46 | 3.38 | 5.07 | 0.3405 | | NN-FT-C | 2.22 | 8.19 | 3.42 | 7.03 | 3.72 | 3.46 | 0.2865 | | NN-FT-N | 1.49 | 8.39 | 3.40 | 2.08 | 3.83 | 1.66 | 0.3526 |
|
|
