A Comparative Study of ARFIMA and Artificial Neural Networks to Forecast Exchange Rate of Dinar Algerian

Abstract

The aim of this study is to compare between (ARFIMA) and (ANN) forecasting models of the exchange rate of the dinar against major currencies in the foreign exchange market, the U.S. dollar, euro, pound sterling using a series of mensuel quotations over the period (2000-2014).

The main finding of this study is that ANN model has better forecasting performance than ARFIMA model for the exchange rate of the dinar Algerien against the U.S. dollar and euro .As for Forecasting the exchange rate of the dinar Algerien against the pound sterling is better with ARFIMA model.