Research Article
[Retracted] Automatic Detection of High-Frequency Oscillations Based on an End-to-End Bi-Branch Neural Network and Clinical Cross-Validation
| Step | Parameters | Setting |
| Signal branch | LSTM hidden size | 100 | LSTM num layers | 2 | Numbers of blocks stacked | [3, 4, 6, 3] |
| TFpic branch | Numbers of blocks stacked | [3, 4, 6, 3] |
| Fusion module | MLP number of hidden layer neurons | 500 |
| Overall | Initial learning rate | 0.01 | Batch size | 32 | Training epochs | 60 |
|
|