Table 2 Hyperparameters used in the experiment
From: Sequence-based drug-target affinity prediction using weighted graph neural networks
Number | Name | Setting |
|---|---|---|
1 | The number of GNN layers | 3 |
2 | The type of GNN pooling layer | Global mean pooling |
3 | Activation function | Sigmoid |
4 | Optimizer | Adam |
5 | Learning rate | 0.001 |
6 | Loss function | Binary cross entropy loss |
7 | Epochs | 1000 |
8 | Batch size | 512 |