Online Selection of Functional Links for Nonlinear System Identification

  • Danilo Comminiello
  • Simone Scardapane
  • Michele Scarpiniti
  • Raffaele Parisi
  • Aurelio Uncini

Abstract

This paper introduces a new method for improving nonlinear modeling performance in online learning by using functional link-based models. The proposed algorithm is capable of selecting the useful nonlinear elements resulting from the functional expansion, while setting to zero the ones that does not bring any improvement of the modeling performance. This allows to reduce any gradient noise due to a possible overestimate of the solution, thus preventing any overfitting phenomena. The proposed model is assessed in several nonlinear identification problems, including different levels of nonlinearity, showing significant improvements.

Keywords

Nonlinear Modeling Functional Links Nonlinear Transformation Nonlinear System Identification Sparse Systems 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Alhamdoosh, M., Wang, D.: Fast decorrelated neural network ensambles with random weights. Inform. Sciences 264, 104–117 (2014)MathSciNetCrossRefGoogle Scholar
  2. 2.
    Comminiello, D., Azpicueta-Ruiz, L.A., Scarpiniti, M., Uncini, A., Arenas-Garcia, J.: Functional link based architectures for nonlinear acoustic echo cancellation. In: Proc. IEEE J. Works. Hands-free Speech Commun. Microph. Arrays (HSCMA 2011), Edinburgh, UK, pp. 180–184 (May 2011)Google Scholar
  3. 3.
    Comminiello, D., Scarpiniti, M., Azpicueta-Ruiz, L.A., Arenas-García, J., Uncini, A.: Functional link adaptive filters for nonlinear acoustic echo cancellation. IEEE Trans. Audio, Speech, Lang. Process. 21(7), 1502–1512 (2013)CrossRefGoogle Scholar
  4. 4.
    Comminiello, D., Scarpiniti, M., Azpicueta-Ruiz, L.A., Arenas-García, J., Uncini, A.: Nonlinear acoustic echo cancellation based on sparse functional link representations. To appear in IEEE Trans. Audio, Speech, Lang. Process. (2014), doi:10.1109/TASLP.2014.2324175Google Scholar
  5. 5.
    Comminiello, D., Scarpiniti, M., Parisi, R., Uncini, A.: Convergence properties of nonlinear functional link adaptive filters. IET Electron. Lett. 49(14), 873–875 (2013)CrossRefGoogle Scholar
  6. 6.
    Doroslovački, M., Deng, H.: Proportionate adaptive algorithms for network echo cancellation. IEEE Trans. Signal Process. 54(5), 1794–1803 (2006)CrossRefGoogle Scholar
  7. 7.
    Duttweiler, D.L.: Proportionate normalized least-mean-squares adaptation in echo cancelers. IEEE Trans. Speech Audio Process. 8(5), 508–518 (2000)CrossRefGoogle Scholar
  8. 8.
    Haykin, S.: Neural Networks and Learning Machines, 3rd edn. Prentice Hall, Upper Saddle River (2008)Google Scholar
  9. 9.
    Karnin, E.: A simple procedure for pruning back-propagation trained neural networks. IEEE Trans. Neural Netw. 1(2), 239–242 (1990)CrossRefGoogle Scholar
  10. 10.
    Naylor, P., Cui, J., Brookes, M.: Adaptive algorithms for sparse echo cancellation. Signal Process. 86(6), 1182–1192 (2006)MATHCrossRefGoogle Scholar
  11. 11.
    Omlin, C., Giles, C.: Pruning recurrent neural networks for improved generalization performance. IEEE Trans. Neural Netw. 5(5), 848–851 (1994)CrossRefGoogle Scholar
  12. 12.
    Orlandi, G., Piazza, F., Uncini, A., Ascone, A.: A biological approach to plasticity in artificial neural netowrks. In: Proc. IEEE Int. J. Conf. Neural Netw (IJCNN 1991), Seattle,WA, July 8-12, vol. 2, pp. 583–586 (1991)Google Scholar
  13. 13.
    Paleologu, C., Benesty, J., Ciochină, S.: Sparse Adaptive Filters for Echo Cancellation. Morgan & Claypool Publishers (2010)Google Scholar
  14. 14.
    Pao, Y.H.: Adaptive Pattern Recognition and Neural Networks. Addison-Wesley, Reading (1989)MATHGoogle Scholar
  15. 15.
    Patra, J.C., Pal, R.N., Chatterji, B.N., Panda, G.: Identification of nonlinear dynamic systems using functional link artificial neural networks. IEEE Trans. Syst., Man, Cybern., B: Cybern. 29(2), 254–262 (1999)CrossRefGoogle Scholar
  16. 16.
    Scardapane, S., Nocco, G., Comminiello, D., Scarpiniti, M., Uncini, A.: An effective criterion for pruning reservoir’s connections in echo state networks. In: Proc. IEEE WCCI - Int. J. Conf. Neural Netw. (IJCNN 2014), Beijing, China, July 6-11 (2014)Google Scholar
  17. 17.
    Scarpiniti, M., Comminiello, D., Parisi, R., Uncini, A.: A collaborative approach to adaptive noise cancellation. In: Proc. XXII Italian Works. Neural Netw. (WIRN), Vietri Sul Mare, Italy, May 17-19 (2012)Google Scholar
  18. 18.
    Sicuranza, G., Carini, A.: On a class of nonlinear filters. In: Tabus, M., Egiazarian, K. (eds.) Festschrift in Honor of Jaakko Astola on the Occasion of his 60th Birthday. TICPS, vol. 47, pp. 115–144 (2009)Google Scholar
  19. 19.
    Sicuranza, G.L., Carini, A.: A generalized FLANN filter for nonlinear active noise control. IEEE Trans. Audio, Speech, Lang. Process. 19(8), 2412–2417 (2011)CrossRefGoogle Scholar
  20. 20.
    Uncini, A.: Fundamentals of Adaptive Signal Processing. Springer (2014) ISBN 978-3-319-02806-4Google Scholar

Copyright information

© Springer International Publishing Switzerland 2015

Authors and Affiliations

  • Danilo Comminiello
    • 1
  • Simone Scardapane
    • 1
  • Michele Scarpiniti
    • 1
  • Raffaele Parisi
    • 1
  • Aurelio Uncini
    • 1
  1. 1.Department of Information Engineering, Electronics and Telecommunications (DIET)“Sapienza” University of RomeRomeItaly

Personalised recommendations