TABLE 2. Different Parameter Values Evaluated for CNN Optimization.

Parameters	Values/Types
Number of convolution layers	1, 2, 3, 4
Number of pooling layers	1, 2, 3
Number of fully connected layers	1, 2, 3
Number of convolutional filters	30, 25, 25, 50, 50, 100,100
Dimension of convolution filters	, , ,
Activation functions	ReLu, softmax
Dropout probability	0.75
Maximum number of iterations	10000
Learning rate