Australian Journal of Applied Mathematics ISSN: 2203-9478 (Print) ISSN: 2203-9486 (Online) Australian Society for Commerce Industry & Engineering www.scie.org.au 1 A Bilayer Feed – Forward Artificial Neural Network for Exchange Rate Prediction Joseph Ackora-Prah Department of Mathematics, Kwame Nkrumah University of Science and Technology Kumasi, Ghana. Tel: +233 208 180 547 Email: ackoraprah@yahoo.co.uk Adu Sakyi Department of Mathematics, Kwame Nkrumah University of Science and Technology Kumasi, Ghana. Tel: +233 245 246 872 Email: ericadusakyi@ymail.com Yao Elikem Ayekple Department of Mathematics, Kwame Nkrumah University of Science and Technology Kumasi, Ghana. Tel: +233 244 442 569 Email: jegyaooo@gmail.com Daniel Gyamfi Department of Mathematics, Kwame Nkrumah University of Science and Technology Kumasi, Ghana. Tel: +233 205 236 649 Email: quamegyamfi@yahoo.co.uk Robert Kofi Acquah. Department of Mathematics, Kwame Nkrumah University of Science and Technology Kumasi, Ghana. Tel: +233 246 445 430 Email: racquah@nims.edu.gh Abstract A feed-forward Neural Network is an interconnection of perceptrons in which data and computations flow in a single direction from the input data to the outputs. We used a two layer feed-forward network using Levenberg – Marquardt Back propagation Neural Network (LMBNN) to forecast the Ghanaian Cedi – US Dollar rate with Treasury bill rates, money supply, consumer price index and inflation. The results were measured with the Mean Squared Error (MSE), Root Mean Squared Error (RMSE) and the Weighted Absolute Percentage Error (WAPE). After careful and extensive training, validation and testing, the Artificial Neural Network (ANN) produced MSE, RMSE, WAPE and an R-value of 0.0010, 0.0324, 2.30%, 0.99634 respectively with a prediction accuracy of 97.70%. Keywords: Exchange Rates, Levenberg – Marquardt, Neural Networks. 1. Introduction There has been an extensive research on the Exchange rate markets. Many believed a random walk model best explained the exchange rate behavior whilst others are of the view that models based on economic factors were successful in predicting exchange rate behavior [Meese and Rogoff (1983)]. Vincenzo Pacelli et al (2010) in their paper concluded that exchange rates can be predicted using artificial intelligence. Most basic needs in life involves the usage of cash. It matters a lot when a person‟s budget is limited because of a particular currency weakness. Exchange rates serve a variety of purposes in the global business world including translation and conversion of foreign currency, expedition of global commerce, the flow of products and services internationally and also serve as economic indicators. According to Lowery (2008), exchange rates are influenced by factors like demand and supply of money which are difficult to control. Productivity, equity flow, hedging activities, interest rate differentials, inflation, Gross Domestic Product (GDP), Current Account Balances (CAB) and