Title: Forecasting exchange rate of Kina against AUD using artificial neural network and time series models

Authors: Mohammad G.M. Khan; Shamsuddin Ahmed; Biman Prasad

Addresses: School of Computing, Information and Mathematical Sciences, The University of the South Pacific, Suva, Fiji ' College of Engineering, Islamic University Madinah, Medina 42351, Saudi Arabia ' 124 Princess Road, Tamavua, Fiji

Abstract: In this paper we propose an Artificial Neural Network model for forecasting exchange rate of Kina against AUD. We use daily exchange rate data during the period of 2 January 2008 to 30 November 2012. The proposed model is compared with autoregressive time series model, exponential smoothing with trend; and Holt-Winter multiplicative and adaptive models. The performance of the models was measured by using various error functions such as root square mean error, mean absolute error, mean absolute deviation, and mean absolute percentage error. The results reveal that the proposed ANN model is an efficient tool for forecasting Kina against AUD more accurately.

Keywords: exchange rate forecasting; exchange rates; artificial neural networks; ANNs; autoregressive time series; exponential smoothing; Holt-Winter multiplicative; adaptive models; Kina; AUD; time series modelling.

DOI: 10.1504/IJBFMI.2015.075342

International Journal of Business Forecasting and Marketing Intelligence, 2015 Vol.2 No.1, pp.37 - 54

Received: 18 Jun 2015
Accepted: 30 Sep 2015
Published online: 15 Mar 2016 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article