Environments Conducive to Evolution of Modularity

  • Vineet R. Khare
  • Bernhard Sendhoff
  • Xin Yao
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4193)


Modularity has been recognised as one of the crucial aspects of natural complex systems. Since these are results of evolution, it has been argued that modular systems must have selective advantages over their monolithic counterparts. Simulation results with artificial neuro-evolutionary complex systems, however, are indecisive in this regard. It has been shown that advantages of modularity, if judged on a static task, in these systems are very much dependent on various factors involved in the training of these systems. We present a couple of dynamic environments and argue that environments like these might be partly responsible for the evolution of modular systems. These environments allow for a better, more direct use of structural information present within modular systems hence limit the influence of other factors. We support these arguments with the help of a co-evolutionary model and a fitness measure based on system performance in these dynamic environments.


Dynamic Environment Hide Unit Related Task Modular System Cross Entropy 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Hrycej, T.: Structure of the Brain. In: Modular Learning in Neural Networks. A Modularized Appproach to Neural Network Classification, pp. 59–82. Wiley, New York (1992)Google Scholar
  2. 2.
    Wagner, G.P., Mezey, J., Calabretta, R.: Natural Selection and the Origin of Modules. In: Callebaut, W., Rasskin-Gutman, D. (eds.) Modularity. Understanding the Development and Evolution of Natural Complex Systems, pp. 33–49. The MIT Press, Cambridge (2005)Google Scholar
  3. 3.
    Ferdinando, A.D., Calabretta, R., Parisi, D.: Evolving Modular Architectures for Neural Networks. In: French, R.M., Sougné, J.P. (eds.) Proceedings of the Sixth Neural Computation and Psychology Workshop, Liege, Belgium, pp. 253–264. Springer, Heidelberg (2001)Google Scholar
  4. 4.
    Bullinaria, J.A.: To Modularize or Not To Modularize? In: Bullinaria, J. (ed.) Proceedings of the 2002 U.K. Workshop on Computational Intelligence (UKCI 2002), Birmingham, pp. 3–10 (2002)Google Scholar
  5. 5.
    Khare, V.R., Yao, X., Sendhoff, B.: Multi-network Evolutionary Systems and Automatic Decomposition of Complex Problems. International Journal of General systems (2006); special issue on Analysis and Control of Complex Systems (to appear)Google Scholar
  6. 6.
    French, R.M.: Catastrophic Interference in Connectionist Networks: Can It Be Predicted, Can It Be Prevented? In: Cowan, J.D., Tesauro, G., Alspector, J. (eds.) Advances in Neural Information Processing Systems, vol. 6, pp. 1176–1177. Morgan Kaufmann, San Francisco (1994)Google Scholar
  7. 7.
    Igel, C., Hüsken, M.: Empirical Evaluation of the Improved Rprop Learning Algorithm. Neurocomputing 50(C), 105–123 (2003)MATHCrossRefGoogle Scholar
  8. 8.
    Thrun, S.B., Mitchell, T.M.: Lifelong Robot Learning. Robotics and Autonomous Systems 15, 25–46 (1995)CrossRefGoogle Scholar
  9. 9.
    Khare, V.R., Yao, X., Sendhoff, B., Jin, Y., Wersing, H.: Co-evolutionary Modular Neural Networks for Automatic Problem Decomposition. In: The 2005 IEEE Congress on Evolutionary Computation, CEC 2005, Edinburgh, Scotland, UK, pp. 2691–2698. IEEE Press, Los Alamitos (2005)CrossRefGoogle Scholar
  10. 10.
    Yao, X., Liu, Y.: A New Evolutionary System for Evolving Artificial Neural Networks. IEEE Transactions on Neural Networks 8(3), 694–713 (1997)CrossRefMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Vineet R. Khare
    • 1
  • Bernhard Sendhoff
    • 2
  • Xin Yao
    • 1
  1. 1.Natural Computation Group, School of Computer ScienceThe University of BirminghamBirminghamUK
  2. 2.Honda Research Institute Europe GmbHOffenbach/MainGermany

Personalised recommendations