Table 5 Final optimized hyperparameter configurations for all compared models.
From: Enhancing groundwater level prediction with a hybrid deep learning model in Jinan City, China
Hyperparameter | LSTM | GRU | STGCN | STGPM |
|---|---|---|---|---|
Number of layers | 2 | 2 | 2 (GCN)/1 (Temporal) | 2 (GraphSAGE)/1 (GRU per branch) |
Hidden dimension | 128 | 128 | 64 (GCN)/128 (Temporal) | 64 (Spatial)/32,64,64 (Temporal) |
Temporal Window | 12 (steps) | 12 (steps) | 12 (steps) | 4, 24, 48 (steps) |
Learning rate | 0.001 | 0.001 | 0.001 | 0.001 |
Batch size | 64 | 64 | 64 | 64 |
Optimizer | AdamW | AdamW | AdamW | AdamW |
Dropout rate | 0.2 | 0.2 | 0.1 | 0.1 |
Early stopping patience | 15 | 15 | 15 | 15 |
