Table 6.

Implementation details of the CNN trained on image data.

Hyper-parameters	Values
Number of convolutional kernels of first layer	64
Number of convolutional kernels of second layer	128
Number of convolutional kernels of third layer	256
Size of convolutional kernels	$3\times 3$
Strides size	2
Input dimension	$100\times 100$
Output dimension	2
Number of convolutional layers	3
Number of fully connected layers	2
Activation function for convolutional and fully connected layers	ReLU
Activation function of last layer	Sigmoid
Adam hyper-parameters	${\beta }_1=0.9,{\beta }_2=0.999$
Learning rate	0.001
Loss function	Binary cross entropy (BCE)
Number of neurons of the fourth layer (fully connected)	256
Number of neurons of fifth layer (fully connected)	128
Dropout probability	0.5
Number of epochs	30
Batch size	128