Forecasting exchange rates using artificial neural networks

Abstract EN

This aims to forecast the exchange rates of the Algerian Dinar against the US dollar using artificial neural networks by building the optimal neural network to know the accuracy of prediction in this method and this is because of the characteristics of exchange rates time series, as they are non-linear, dynamic and random.

The lagged values of the monthly data of the Algerian Dinar exchange rates against the US dollar were used as inputs for three artificial neural networks that differed in their architectures in terms of the number of neurons in their hidden layer, and they were compared in terms of prediction efficiency.

The results of the study showed that the artificial neural network which contains eight hidden neurons, has outperformed the other networks in predicting accuracy.