Table 3.

Deep learning classifier architecture and hyper parameters.

Parameter	ANN	CNN	RNN
Units	32-16-8	512–256	–
Layers	1-2-3	1–2	1
Activation function	ReLU	ReLU	ReLU
Learning rate	1e−3	1e−3	1e−3
Loss function	Binary cross entropy	Binary cross entropy	Binary cross entropy
Epocs	250	250	250
Optimizer	SGD	SGD	SGD
Decay	1e−5	1e−5	1e−5
Momentum	0.3	0.3	0.3
Fully Conn.units	–	2048–1024	2048–1024
Fully Conn.layer	–	1–2	1–2
RNN units	–	–	512
Dropout	–	–	0.25