Table 6 ANN hyperparameters.
Parameter | Values |
|---|---|
Hidden neurons/nodes | 12/12 |
Network type | Feed-forward back-propagation |
Total dataset | 2813 |
Testing | 30% |
Rate of learning | 0.01 |
Training method | Levenberg–Marquardt |
Data division | Random |
Transfer function (Input layer) | PURELIN |
Non-linear parameters | 20 |
Transfer function (Hidden layer) | TANSIG |
