Currency exchange rate prediction and neural network design strategies (original) (raw)

Abstract

This paper describes a non trivial application in forecasting currency exchange rates, and its implementation using a multi-layer perceptron network. We show that with careful network design, the backpropagation learning procedure is an effective way of training neural networks for time series prediction. The choice of squashing function is an important design issue in achieving fast convergence and good generalisation performance. We evaluate the use of symmetric and asymmetric squashing functions in the learning procedure, and show that symmetric functions yield faster convergence and better generalisation performance. We derive analytic results to show the conditions under which symmetric squashing functions yield faster convergence, and to quantify the upper bounds on the convergence improvement. The network is evaluated both for long-term forecasting without feedback (i.e. only the forecast prices are used for the remaining trading days), and for short- term forecasting with hourly feedback. The network learns the training set near perfect, and shows accurate prediction, making at least 22% profit on the last 60 trading days of 1989.

Access this article

Log in via an institution

Subscribe and save

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

References

  1. Kimoto T. et al. Stock market prediction with modular neural networks. In: Proceedings of IJCNN-90, San Diego, CA, 1990
  2. Weigend AS, Huberman BA, Rumelhart DE. Predicting the future: a connectionist approach. Stanford (CA): Stanford Univ Technical Report Stanford-PDP-90-01 (Submitted to Int J Neural Syst 1990)
  3. Denker J, Schwartz D, Wittner B, Solla S, Howard R, Jackel L, Hopfield J. Large automatic learning rule extraction and generalisation. Complex Syst 1987; 1: 877–922
    Google Scholar
  4. Patarnello S, Carnevali P. Learning networks of neurons with boolean logic. Europhysics Lett 1987; 4(4): 503–508
    Google Scholar
  5. le Cun Y. Generalisation and network design strategies. Toronto (Canada): Univ of Toronto, Technical Report CRG-TR-89-4
  6. Refenes AN, Vithlani S. Constructive learning by specialisation., In: Proceedings fo ICANN-91, Helsiniki, Finland, 1991
  7. Refenes AN, Alippi C. Histological image understanding by error backpropagation. Microproc Microprog 1991; 32: 437–446
    Google Scholar
  8. Brown RG. Smoothing, forecasting and prediction of discrete time series, Prentice-Hall, NJ, 1963
    Google Scholar
  9. Mendenhall W et al. Statistics for management and economics, PWS-KENT Publishing, Boston, MA, 1989
    Google Scholar

Download references

Author information

Author notes

  1. S. A. Karoussos
    Present address: Codeware Research Ltd., UK

Authors and Affiliations

  1. Department of Computer Science, University College London, Gower Street, WC1 6BT, London, UK
    A. N. Refenes, M. Azema-Barac & L. Chen
  2. Department of Economics, Queen Mary and Westfield College, Mile End Road, E1 4NS, London, UK
    S. A. Karoussos

Authors

  1. A. N. Refenes
  2. M. Azema-Barac
  3. L. Chen
  4. S. A. Karoussos

Additional information

This research is partially supported by the European Communities ESPRIT Programme under Project FOES, entitled ‘Front Office Environment Study’.

Rights and permissions

About this article

Cite this article

Refenes, A.N., Azema-Barac, M., Chen, L. et al. Currency exchange rate prediction and neural network design strategies.Neural Comput & Applic 1, 46–58 (1993). https://doi.org/10.1007/BF01411374

Download citation

Keywords