Table A1.
Convolutional and long short-term memory (CNN + LSTM) model summary.
| Layer Type | Output Shape | Param # |
|---|---|---|
| Conv1D | (None, 2993, 8) | 136 |
| Activation (ReLU) | (None, 2993, 8) | 0 |
| MaxPooling1D | (None, 997, 8) | 0 |
| Conv1D | (None, 990, 16) | 1040 |
| Activation (ReLU) | (None, 990, 16) | 0 |
| MaxPooling1D | (None, 330, 16) | 0 |
| Conv1D | (None, 323, 32) | 4128 |
| Activation (ReLU) | (None, 323, 32) | 0 |
| MaxPooling1D | (None, 107, 32) | 0 |
| LSTM | (None, 107, 64) | 24,832 |
| LSTM | (None, 64) | 33,024 |
| Dense | (None, 5) | 325 |
| Total Params | 63,485 | |
| Trainable Params | 63,485 | |
| Non-Trainable Params | 0 |