Table 3.
The range of hyperparameters used in the Bayesian optimization process.
Model | Hyperparameter range |
---|---|
ATT_BO | Activation function: (ReLU, Linear) |
LSTM_BO | Activation function: (ReLU, Linear, Tanh) |
Dropout rate: (0.0,0.1,0.2,0.3,0.4,0.50) | |
Number of neurons: (32,64,128,256) | |
CNN_BO | Size of kernel: (2,3,4,5,6) |
Stride: (1,2) | |
Number of neurons: (32,64,128,256) |