Forecasting of Currency Exchange Rate Using Artificial Neural Network: A Case Study of Solomon Island Dollar

Abstract

The use of neural network models for currency exchange rate forecasting has received much attention in recent time. In this paper, we propose an exchange rate forecasting model based on artificial neural network. We tested our model on forecasting the exchange rate of Solomon Islands Dollar against some major trading currencies of the country such as, Australian Dollar, Great Britain Pound, Japanese yen, and Euro. We compared the performance of our model with that of the single exponential smoothing model; the double exponential smoothing with trend model; and Holt-Winter multiplicative and additive seasonal and multiple linear regression model. The performance of the models was measured using the error function, root mean square error (RMSE). The empirical result reveals that the proposed model is more efficient and accurate in forecasting currency exchange rate in comparison to the regression and time series models.