Table 2 GRU parameter setting for BG prediction model.
From: Event-triggered smart dual hormone artificial pancreas for patient-specific drug delivery
S. no | GRU parameter | Value |
|---|---|---|
1. | Time steps | 30s |
2. | Input Layer size | 100 |
3. | Output layer size | 3 |
4. | Hidden layers | 2 |
5. | Dropout rate | 0.2–0.5 |
6. | Epochs | 100 |
7. | Batch size | 64 |
8. | Activation function | ReLU |
9. | Loss function | MAE |
10. | Prediction horizon | Short-term (15–60 min) |
11. | Optimizer | Adam |
12. | Learning rate | 0.001 |
