Table 3 Hyperparameter tuning.
From: Dual smart sensor data-based deep learning network for premature infant hypoglycemia detection
Hyperparameter | Potential Values | Selected Value |
|---|---|---|
GRU Units | 32, 64, 128 | 64 |
LSTM Units | 32, 64, 128 | 64 |
Batch Size | 16, 32, 64 | 32 |
Epochs | 50, 100, 150 | 100 |
Learning Rate | 0.001, 0.0005, 0.0001 | 0.0005 |
Dropout Rate | 0.2, 0.3, 0.5 | 0.3 |
Optimizer | Adam, RMSprop, SGD | Adam |
Activation Function | softmax, sigmoid | softmax |
Validation Split | 0.1, 0.2 | 0.2 |
