Forecasting Nigeria Foreign Exchange Rates Using Artificial Neural Network

Abstract

Artificial neural networks (ANNs) are computing models for information processing and pattern identification. They grow out of research interest in modeling biological neural systems, especially human brains. An ANN is a network of many simp le computing units called neurons or cells, which are highly interconnected and organized in layers. Each neuron performs the simple task of information processing by converting received inputs into processed outputs. In past two decades, ANN has been appl ied in Economics, Finance and other sectors. The foreign exchange market assists international trade and investment by enabling currency conversion. In this study we applied a time delayed neural network model for foreca s ting dail y foreign exchange rate of a US Dollar to Naira for Nigeria by using Artificial Neural Network (ANN) methodology on the basis of daily data for September 2011 t o February 2012. We compared ANN with Single Exponential Smoothening (SES) and Autoregressive - Integrated - Moving - Average (A RIMA) models, the ANN forecasting tool proved to be more accurate than the SES and ARIMA as it had a smaller root mean squared error of 0.6995 as compared to the root mean squared error of the SES which was 0.9890 and ARIMA which was 0.7880. More research work can be carried out by comparing ANN with other available forecasting tools. Key words: Artificial Neural Networks, Forecasting, Foreign Exchange Rate, Single Exponential Smoothening