Brent Crude Oil Price Forecast Utilizing Deep Neural Network Architectures

Table 4

Results of Bi-LSTM with SGDM solver.

Parameter	Value
Root-mean-square error (RMSE)	2.18
Maximum error	7.74
Modeling time	822
Mean squared error (MSE)	3.60
Feedback regressors	[1 2 3 4 ]
Hidden layers	100