Table 5 Optimization of ANN model parameters.
From: Deep learning time-series modeling for assessing land subsidence under reduced groundwater use
Initial value | Stopping criteria | Target value | |
|---|---|---|---|
Epoch | 0 | 9 | 103 |
Elapsed time | – | 1.66 s | – |
Performance | 1.09 \(\times \hspace{0.17em}\)10–3 | 6.6\(0\hspace{0.17em}\times \hspace{0.17em}\)10–3 | 0 |
Gradient | 6.00 × 10–1 | 1.7\(0\hspace{0.17em}\times \hspace{0.17em}\)10–2 | 10–7 |
mu | 10–3 | 10–4 | 1010 |
Validation checks | 0 | 6 | 6 |
Loss function | MSE | Activation function | Hyperbolic tangent function |
Number of hidden layer | 20 |
