Table 3: The list of LSTM network parameters.

Model parameters LSTM

Maximum iterations6.40 + 04
Weights initialization
Truncated BPTT length120
Batch training samples30
Initial learning rate2.00 − 03
Dropout probability0.5
Gradient regularization factor10
Activation functionTanh
Optimization algorithmAdam
Propagation direction of time seriesOne-way
Hidden layers2
Hidden nodes185