Table 6 Network structure parameters.
Layer | Nodes | Activation | Description |
|---|---|---|---|
Input | State vector size | - | Accepts the state vector \(\:{s}_{i}^{t}\) |
Hidden 1 | 256 | ReLU | Learns abstract representations of the input state |
Hidden 2 | 128 | ReLU | Further processes the learned features |
Hidden 3 | 64 | ReLU | Refines the learned representations |
Output | Action vector size | tanh | Generates the action vector \(\:{a}_{i}^{t}\) |
