Advertisement

Discriminant Independent Component Analysis

  • Chandra Shekhar Dhir
  • Soo Young Lee
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5788)

Abstract

In this paper, a new approach for extraction of discriminative and independent features is proposed. The proposed discriminant ICA (dICA) method jointly maximizes the inter-class variance and Negentropy of a given feature. Experimental results shows much improved classification performance when dICA features are used for recognition tasks over conventional ICA features. Moreover, dICA features show higher Fisher criterion score value suggesting a better capability to do class discrimination.

Keywords

Independent Component Analysis Independent Component Analysis Independent Feature Unsupervised Feature Extraction Face Image Database 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Hinton, G., Sejnowski, T.: Unsupervised learning: Foundations of Nuerla Computation. MIT Press, Cambridge (1999)Google Scholar
  2. 2.
    Turk, M., Pentland, A.: Eigenfaces for recognition. J. Cognitive Neuroscience 3, 71–86 (1991)CrossRefGoogle Scholar
  3. 3.
    Dhir, C.S., Lee, S.Y.: Hybrid Feature Selection: Combining Fisher Criterion and Mutual Information for Efficient Feature Selection. In: Köppen, M., Kasabov, N., Coghill, G. (eds.) ICONIP 2008. LNCS, vol. 5506, pp. 613–620. Springer, Heidelberg (2009)Google Scholar
  4. 4.
    Hyvarinen, A.: Fast and Robust Fixed-Point Algorithms for Independent Component Analysis. IEEE Trans. on Neural Networks 10, 626–634 (1999)CrossRefGoogle Scholar
  5. 5.
    Hyvarinen, A., et al.: Independent Component Analysis. John Wiley & sons, Inc., Chichester (2001)CrossRefzbMATHGoogle Scholar
  6. 6.
    Lee, D.D., Seung, H.S.: Learning the parts of objects by non-negative matrix factorization. Nature 401, 788–791 (1999)CrossRefGoogle Scholar
  7. 7.
    Bell, A.J., Sejnowski, T.J.: An information-maximization approach to blind separation and blind deconvolution. Neural Computation 7, 1129–1159 (1995)CrossRefGoogle Scholar
  8. 8.
    Bartlett, M.S., et al.: Face recognition by independent component analysis. IEEE Transactions on Neural Networks 13(6), 1450–1462 (2002)CrossRefGoogle Scholar
  9. 9.
    Wang, G., et al.: Feature selection with conditional mutual information MaxMin in text categorization. In: Proc. Int. Conf. on Information and Knowledge Management, pp. 342–349 (2004)Google Scholar
  10. 10.
    Su, H., et al.: Face Recognition Method Using Mutual Information and Hybrid Feature. In: Proc. Int. Conf. on Computational Intelligence and Multimedia Applications, pp. 436–440 (2003)Google Scholar
  11. 11.
    Kaynak, C.: Methods of Combining Multiple Classifiers and Their Applications to Handwritten Digit Recognition MSc Thesis, Institute of Graduate Studies in Science and Engineering, Bogazici University (1995)Google Scholar
  12. 12.
    IMDB: Asian face image database. Technical report POSTECH, Korea, http://nova.postech.ac.kr/
  13. 13.
    Goldberger, J., Roweis, S., Hinton, G., Salakhutdinov, R.: Neighbourhood components analysis. In: Advances in Neural Information Processing Systems, vol. 17, pp. 513–520. MIT Press, Cambridge (2005)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Chandra Shekhar Dhir
    • 1
    • 3
  • Soo Young Lee
    • 1
    • 2
    • 3
    • 4
  1. 1.Department of Bio and Brain EngineeringKorea
  2. 2.Department of Electrical Engineering and Computer ScienceKorea
  3. 3.Brain Science Research Center, KAISTDaejeonKorea
  4. 4.RIKEN Brain Science Institute, Wako-shiSaitamaJapan

Personalised recommendations