Table 1.

Summary of the different training parameters used in this study.

Network	Model architecture	Batch size	No.of epochs
GRU RNN	2x 100 GRU - 100 Dense (ReLU) - 10 Softmax	128	200
LSTM RNN	100 Dense (SELU) - 2x 100 LSTM - 10 Dense	128	200
Phased LSTM	2x 250 Phased LSTM - 10 Dense	16	50

Tha Adam optimizer with a learning of 0.001 was used for all the networks.