Adaptive Mechanisms for Classification Problems with Drifting Data

  • Zoheir Sahel
  • Abdelhamid Bouchachia
  • Bogdan Gabrys
  • Paul Rogers
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4693)


Most work on supervised learning is undertaken on static problems. However, in many real world classification problems, the environment in which the classifiers operate is dynamic i.e. the descriptions of classes change with time. In this paper, the process of generating drifting data is introduced in order to assess two adaptive approaches that deal with dynamically changing data. These approaches are: retraining on evolving data set and incremental learning. The empirical evaluation has shown that both these approaches improve the performance compared to the non-adaptive mode though a number of outstanding research issues remain.


Concept drift adaptive classification Incremental learning 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Kelly, M.G., Hand, D.J., Adams, N.M.: The impact of changing populations on classifier performance. In: Proc of the fifth ACM SIGKDD international conference on Knowledge discovery and data mining, pp. 367–371 (1999)Google Scholar
  2. 2.
    Gabrys, B.: Learning hybrid neuro-fuzzy classifier models from data: to combine or not to combine? Fuzzy Sets and Systems 147(1), 39–56 (2004)MathSciNetCrossRefzbMATHGoogle Scholar
  3. 3.
    Widmer, G., Kubat, M.: Learning in the presence of concept drift and hidden contexts. Machine Learning 23(1), 69–101 (1996)Google Scholar
  4. 4.
    Salganicoff, M.: Density-adaptive learning and forgetting. In: Proceedings of the Tenth International Conference on Machine Learning, pp. 276–283 (1993)Google Scholar
  5. 5.
    Abraham, W.C., Robins, A.: Memory retention-the synaptic stability versus plasticity dilemma. Trends in Neurosciences 28(2), 73–78 (2005)CrossRefGoogle Scholar
  6. 6.
    French, R.M.: Pseudo-recurrent connectionist networks: An approach to the Sensitivity-Stability dilemma. Connection Science 9(4), 353–380 (1997)CrossRefGoogle Scholar
  7. 7.
    Lim, C.P., Harrison, R.F.: Online pattern classification with multiple neural network systems: an experimental study. Systems, Man and Cybernetics, Part C, IEEE Transactions 33(2), 235–247 (2003)Google Scholar
  8. 8.
    Bouchachia, A., Mittermeir, R.: Towards incremental fuzzy classifiers. Soft Computing 11(2), 193–207 (2007)CrossRefGoogle Scholar
  9. 9.
    Bouchachia, A., Gabrys, B., Sahel, Z.: Overview of some incremental learning algorithms (2007)Google Scholar
  10. 10.
    Hettich, S., Blake, C., Merz, C.: UCI repository of machine learning databases (1998),
  11. 11.
    Duin, R.P.W.: PRTools 3 Manual, provided with PRTools 3.1.6 Toolkit (2000)Google Scholar
  12. 12.
    Grossberg, S.: Nonlinear neural networks: principles, mechanism, and architectures. Neural Networks 1, 17–61 (1988)CrossRefGoogle Scholar
  13. 13.
    Carpenter, G., Grossberg, S., Rosen, D.: Fuzzy ART: Fast stable learning and categorization of analog patterns by an adaptive resonance system. Neural Networks 4(6), 759–771 (1991)CrossRefGoogle Scholar
  14. 14.
    Salzberg, S.: A nearest hyperrectangle learning method. Machine learning 6, 277–309 (1991)Google Scholar
  15. 15.
    Gabrys, B., Bargiela, A.: General fuzzy min-max neural network for clustering and classification. IEEE Transactions on Neural Networks 11(3), 769–783 (2000)CrossRefGoogle Scholar
  16. 16.
    Fritzke, B.: A growing neural gas network learns topologies. In: Advances in neural information processing systems, pp. 625–632 (1995)Google Scholar
  17. 17.
    Bouchachia, A.: Incremental learning via function decomposition. In: Proceedings of the 5th ICMLA, pp. 63–68. IEEE Computer Society, Los Alamitos (2006)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2007

Authors and Affiliations

  • Zoheir Sahel
    • 1
  • Abdelhamid Bouchachia
    • 2
  • Bogdan Gabrys
    • 1
  • Paul Rogers
    • 1
  1. 1.School of Design, Engineering and Computing, Bournemouth UniversityUK
  2. 2.Dept. of Informatics, University of KlagenfurtAustria

Personalised recommendations