From: DeepSuccinylSite: a deep learning based approach for protein succinylation site prediction
Parameters | Settings |
---|---|
Embedding Output Dimension | 21 |
Learning Rate | 0.001 |
Batch Size | 256 |
Epochs | 80 |
Conv2d_1 number of filters | 64 |
Conv2d_1 filter size | 17 × 3 (For window size 33) |
Conv2d_1 padding | Disabled |
Dropout | 0.6 |
Conv2d_1 number of filters | 128 |
Conv2d_1 filter size | 3 × 3 |
Conv2d_1 padding | Enabled |
Dropout | 0.6 |
MaxPooling2d | 2 × 2 |
Dense 1 | 768 |
Dropout | 0.5 |
Dense_2 | 256 |
Dropout | 0.5 |
Checkpointer | Best validation accuracy |