Table 1 Deep neural network hyperparameter tuning and the range for each hyperparameter
Model parameters | Value |
|---|---|
Units | From 50 to 150 by a step of 10 |
Epochs | 10 |
Hidden layers | 3, 4, 5, and 6 |
Learning rate | 0.005 |
Loss function | Automatic setting |
Regularization method | Dropout |
Activation function | Rectified linear activation unit function |