A Review of Artificial Neural Networks: How Well Do They Perform in Forecasting Time Series?
At the beginning of the 90’s, Artificial Neural Networks (ANNs) started their applications in finance. The ANNs are data-drive, self-adaptive and non-linear methods that do not require specific assumptions about the underlying model. In general, there are five groups of networks used as forecasting tools: 1) Feedforward Networks, like the Multilayer Perceptron (MLP), 2) Recurrent Networks, 3) Polynomial Networks, 4) Modular Networks, and 5) Support Vector Machine. This paper carries out a review of the specialized literature on ANNs and makes a comparative analysis according to their performance in forecasting stock indices and exchange rates. The objective is to assess the performance when applying different types of networks in relation to MLP. It is shown that the MLP is the best network in forecasting time series. However, it is shown that the MLP has important delimitations in several respects: network architecture, basic functions and initialization weights.
References listed on IDEAS
Please report citation or reference errors to , or , if you are the registered author of the cited work, log in to your RePEc Author Service profile, click on "citations" and make appropriate adjustments.:
- Marcel Scharth & Marcelo Cunha Medeiros, 2006.
"Asymmetric effects and long memory in the volatility of Dow Jones stocks,"
Textos para discussão
532, Department of Economics PUC-Rio (Brazil).
- Scharth, Marcel & Medeiros, Marcelo C., 2009. "Asymmetric effects and long memory in the volatility of Dow Jones stocks," International Journal of Forecasting, Elsevier, vol. 25(2), pages 304-327.
- Hamid, Shaikh A. & Iqbal, Zahid, 2004. "Using neural networks for forecasting volatility of S&P 500 Index futures prices," Journal of Business Research, Elsevier, vol. 57(10), pages 1116-1125, October.
- Perez-Rodriguez, Jorge V. & Torra, Salvador & Andrada-Felix, Julian, 2005. "STAR and ANN models: forecasting performance on the Spanish "Ibex-35" stock index," Journal of Empirical Finance, Elsevier, vol. 12(3), pages 490-509, June.
- Adeodato, Paulo J.L. & Arnaud, Adrian L. & Vasconcelos, Germano C. & Cunha, Rodrigo C.L.V. & Monteiro, Domingos S.M.P., 2011. "MLP ensembles improve long term prediction accuracy over single networks," International Journal of Forecasting, Elsevier, vol. 27(3), pages 661-671, July.
- Leung, Mark T. & Daouk, Hazem & Chen, An-Sing, 2000. "Forecasting stock indices: a comparison of classification and level estimation models," International Journal of Forecasting, Elsevier, vol. 16(2), pages 173-190.
- Hutchinson, James M & Lo, Andrew W & Poggio, Tomaso, 1994.
" A Nonparametric Approach to Pricing and Hedging Derivative Securities via Learning Networks,"
Journal of Finance,
American Finance Association, vol. 49(3), pages 851-89, July.
- James M. Hutchinson & Andrew W. Lo & Tomaso Poggio, 1994. "A Nonparametric Approach to Pricing and Hedging Derivative Securities Via Learning Networks," NBER Working Papers 4718, National Bureau of Economic Research, Inc.
- Kuan, Chung-Ming & Liu, Tung, 1995. "Forecasting Exchange Rates Using Feedforward and Recurrent Neural Networks," Journal of Applied Econometrics, John Wiley & Sons, Ltd., vol. 10(4), pages 347-64, Oct.-Dec..
- Khurshid Kiani & Terry Kastens, 2008. "Testing Forecast Accuracy of Foreign Exchange Rates: Predictions from Feed Forward and Various Recurrent Neural Network Architectures," Computational Economics, Society for Computational Economics, vol. 32(4), pages 383-406, November.
When requesting a correction, please mention this item's handle: RePEc:inp:inpana:v:6:y:2013:i:2:p:7-15. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: ()
If references are entirely missing, you can add them using this form.