Table 2. Details of CNN module and prediction module.
| Layers | Details | |
|---|---|---|
| Multi-layer CNN | Conventional layer(+ReLU) | 32 filters |
| Conventional layer(+ReLU) | 64 filters | |
| Conventional layer(+ReLU) | 128 filters | |
| Dropout | P = 0.3 | |
| MaxPooling | pool_size = 2 | |
| Multi-layer DNN | Fully connected layer(+ReLU) | 128 neurons |
| Fully connected layer(+ReLU) | 64 neurons | |
| Fully connected layer(+ReLU) | 32 neurons | |
| Dropout | P = 0.3 | |
| Output layer | Fully connected layer (+softmax) | 2 neurons |