Table 2 Network structure.
From: A fiber channel modeling method based on complex neural networks
Parameter | Network | |
|---|---|---|
C-GNet | C-DNet | |
Input layer dimension | 41 | 40 |
Hidden layer | (144, 128, 32)a | (128, 128, 32) |
Output layer | 4 | 1 |
Learning rate | 0.001 | 0.001 |
Activation function | \(\mathbb {C}\)ReLU, the last layer does not use the activation function | \(\mathbb {C}\)ReLU, the last layer uses \(\mathbb {C}\)Sigmoid |
Optimizer | Adam | Adam |
Normalization | Kaiming initialization and \(\mathbb {C}\)BN | Kaiming initialization and \(\mathbb {C}\)BN |
