Comparison of Performance of Different Functions in Functional Link Artificial Neural Network: A Case Study on Stock Index Forecasting

Conference paper
Part of the Smart Innovation, Systems and Technologies book series (SIST, volume 31)


The rapid growth of world economy and globalization has been attracting researchers to develop intelligent forecasting models for stock market prediction. In order to forecasting the stock market trend efficiently, this paper developed four single layer low complex forecasting models known as functional link artificial neural network (FLANN). Different basis functions such as Trigonometric, Chebysheb, Legendre and Lagurre polynomials are used for functional expansion of input signals to achieve higher dimensionality. The models are termed as TFLANN, CFLANN, LeFLANN and LFLANN respectively. The weight and bias vectors are optimized by genetic algorithm (GA). The number of functional expansion for each models are optimized by GA during the training process instead of fixing it earlier, which is the novelty of this research work. The models are employed to forecast the one-day-ahead prediction of three fast growing global stock markets. Different types of FLANN are considered and their comparative performance is investigated.


Stock market forecasting Functional link artificial neural network Genetic algorithm Chebysheb polynomial Legendre polynomial 


  1. 1.
    Kwon, Y.K., Moon, B.R.: A hybrid neuro-genetic approach for stock forecasting. IEEE Trans. Neural Networks 18(3), 851–864 (2007)CrossRefMathSciNetGoogle Scholar
  2. 2.
    Guangxu, Z.: RBF based time-series forecasting. J. Comput. Appl. 9, 2179–2183 (2005)Google Scholar
  3. 3.
    Yu, L., Zhang, Y. Q.: Evolutionary fuzzy neural networks for hybrid financial prediction. IEEE Trans. Syst. Man Cybern.—Part C Appl. Rev. 35(2), 244–249 (2005)Google Scholar
  4. 4.
    Nayak, S.C., Misra, B.B., Behera, H.S.: Index prediction using neuro-genetic hybrid networks: a comparative analysis of perfermance. International Conference on Computing Communication and Application, pp. 22–24. IEEE (2012). doi: 10.1109/ICCCA.2012.6179215 Google Scholar
  5. 5.
    Nayak, S.C., Misra, B.B., Behera, H.S.: Stock index prediction with neuro-genetic hybrid techniques. Int. J. Comput. Sci. Inform. 2, 27–34 (2012)Google Scholar
  6. 6.
    Pao, Y.H.: Adaptive pattern recognition and neural networks. Addison-Wesley, Boston (1989)Google Scholar
  7. 7.
    Pao, Y.H., Takefuji, Y.: Functional-link net computing: theory, system architecture, and functionalities. Computer 25, 76–79 (1992)CrossRefGoogle Scholar
  8. 8.
    Patra, J.C., Bos, A.V.D.: Modeling of an intelligent pressure sensor using functional link artificial neural networks. ISA Trans. Elsevier 39, 15–27 (2000)CrossRefGoogle Scholar
  9. 9.
    Majhi, R., Panda, G., Sahoo, G.: Development and performance evaluation of FLANN based model for forecasting of stock markets. Expert Syst. Appl. 36, 6800–6808 (2009)CrossRefGoogle Scholar
  10. 10.
    Patra, J.C., Pal, R.N., Baliarsingh, R., Panda, G.: Nonlinear channel equilization for QAM signal constellation using artificial neural network. IEEE Trans. Syst. Man Cybern. Part B 29(2), 262–271 (1999)CrossRefGoogle Scholar
  11. 11.
    Patra, J.C., Kim, W., Meher, P.K., Ang, E.L.: Financial Prediction of Major Indices Using Computational Efficient Artificial Neural Networks, pp. 2114–2120. IJCNN, Vancouver (2006)Google Scholar
  12. 12.
    Nayak, S.C., Misra, B.B., Behera, H.S.: Impact of data normalization on stock index forecasting. Int. J. Comp. Inf. Syst. Ind. Manag. Appl. 6, 357–369 (2014)Google Scholar

Copyright information

© Springer India 2015

Authors and Affiliations

  1. 1.Veer Surendra Sai University of TechnologyBurlaIndia
  2. 2.Silicon Institute of TechnologyBhubaneswarIndia

Personalised recommendations