Table 3.

LSTM hyper parameters.

Number of layers	5
Neurons in each layer	128
Optimizer	ADAM
Dropout	0.2
Activation Layer	Softmax
Epoch	300
Batch size	32
Training data	75%
Testing data	25%