Back to Article

Nigeria Exchange Rate Volatility: A Comparative Study of Recurrent Neural Network LSTM and Exponential Generalized Autoregressive Conditional Heteroskedasticity Models

Journal of Artificial Intelligence and Big Data | Vol 4, Issue 2

Table 4. RNN-LSTM AIC Result for Train and Validation Result for Euro, Pound Sterling and US Dollars

EuroPound SterlingUS Dollars
Structure of the LSTM ModelAIC of TrainingAIC of TestingAIC of TrainingAIC of TestingAIC of TrainingAIC of Testing
p = 1, q = 15.445.655.565.644.324.64
p = 5, q = 13.563.713.883.922.762.97
p = 10, q = 13.123.223.403.822.222.67
p = 20 q = 13.323.393.333.912.512.74
p = 40, q = 13.053.493.463.832.352.53
p = 80, q = 17.218.129.329.932.692.94
p = 4, q = 46.636.927.757.925.315.59
p = 8, q = 86.927.117.878.025.425.88
p = 4, q = 406.646.937.547.746.427.03
p = 8, q = 805.315.666.116.635.876.34