Table 1.

Hyper parameter and structure of vanilla LSTM, stacked LSTM, ED-LSTM, BiLSTM models.

Models	Layers	Number	Bias	Models	Layers	Number	Bias
		of units	regularizer			of units	regularizer
			L1				L1
Vanilla LSTM	LSTM	200	0.02	BiLSTM	Bidirectional	250	0.4
	Dropout	0.2	–		Dropout	0.2	–
	Dense	1	–		Dense	1	–
Stacked LSTM	LSTM	130	0.04	ED_LSTM	LSTM	125	0.02
	Dropout	0.4	–		Dropout	0.2	–
	LSTM	100	0.04		Repeat Vector	4	–
	Dropout	0.4	–		LSTM	75	0.02
	LSTM	75	0.04		Dropout	0.2	–
	Dense	1	–		Dense	1	–