Table 4.

Summary of the hyperparameter settings for our AE model.

Hyperparameters	Values
Epochs	100
Batch size	32
Initial learning rate	0.001
Dropout	0.2
L1 regularization	0.0001
Max-pooling size	2
Up-sampling size	2
Filter size	5
Number of filters	(32, 64, 128, 64, 32, 4)
Strides	1
Padding	Same
Optimizer	Adam
Activation function	ReLU except the output layer (Softmax)