The Introduction of Time-Scales in Reservoir Computing, Applied to Isolated Digits Recognition

  • Benjamin Schrauwen
  • Jeroen Defour
  • David Verstraeten
  • Jan Van Campenhout
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4668)

Abstract

Reservoir Computing (RC) is a recent research area, in which a untrained recurrent network of nodes is used for the recognition of temporal patterns. Contrary to Recurrent Neural Networks (RNN), where the weights of the connections between the nodes are trained, only a linear output layer is trained. We will introduce three different time-scales and show that the performance and computational complexity are highly dependent on these time-scales. This is demonstrated on an isolated spoken digits task.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Omlin, C.W., Giles, C.L.: Constructing deterministic finite-state automata in sparse recurrent neural networks. In: IEEE International Conference on Neural Networks (ICNN’94), Piscataway, NJ, pp. 1732–1737. IEEE Computer Society Press, Los Alamitos (1994)Google Scholar
  2. 2.
    Kilian, J., Siegelmann, H.T.: The dynamic universality of sigmoidal neural networks. Information and Computation 128, 48–56 (1996)MATHCrossRefGoogle Scholar
  3. 3.
    Hammer, B., Steil, J.J.: Perspectives on learning with recurrent neural networks. In: Proceedings of ESANN (2002)Google Scholar
  4. 4.
    Maass, W., Natschläger, T., Markram, H.: Real-time computing without stable states: A new framework for neural computation based on perturbations. Neural computation 14(11), 2531–2560 (2002)MATHCrossRefGoogle Scholar
  5. 5.
    Jaeger, H.: Short term memory in echo state networks. Technical Report GMD Report 152, German National Research Center for Information Technology (2001)Google Scholar
  6. 6.
    Jaeger, H., Lukosevicius, M., Popovici, D.: Optimization and applications of echo state networks with leaky integrator neurons. Neural Networks (to appear, 2007)Google Scholar
  7. 7.
    Verstraeten, D., Schrauwen, B., D’Haene, M., Stroobandt, D.: A unifying comparison of reservoir computing methods. Neural Networks (to appear, 2007)Google Scholar
  8. 8.
    Lyon, R.: A computational model of filtering, detection and compression in the cochlea. In: Proceedings of the IEEE ICASSP, pp. 1282–1285. IEEE Computer Society Press, Los Alamitos (1982)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Authors and Affiliations

  • Benjamin Schrauwen
    • 1
  • Jeroen Defour
    • 1
  • David Verstraeten
    • 1
  • Jan Van Campenhout
    • 1
  1. 1.Electronics and Information Systems Department, Ghent UniversityBelgium

Personalised recommendations