Table 4.
Neurons of two NN models.
| Input layer: 6 neurons, including six form elements of PDAs | |
| NN-FE model | Hidden layer: 4 neurons, (6 + 1)/2 = 3.5 ≒ 4 |
| Output layer: 1 neuron for the S-C image value | |
|
| |
| Input layer: 19 neurons, including 19 types of six form elements | |
| NN-FT model | Hidden layer: 10 neurons, (19 + 1)/2 = 10 |
| Output layer: 1 neuron for the S-C image value | |