MULTI-CLASS FROM BINARY - Divide to conquer

Anderson Rocha, Siome Goldenstein

2009

Abstract

Several researchers have proposed effective approaches for binary classification in the last years. We can easily extend some of those techniques to multi-class. Notwithstanding, some other powerful classifiers (e.g., SVMs) are hard to extend to multi-class. In such cases, the usual approach is to reduce the multi-class problem complexity into simpler binary classification problems (divide-and-conquer). In this paper, we address the multi-class problem by introducing the concept of affine relations among binary classifiers (dichotomies), and present a principled way to find groups of high correlated base learners. Finally, we devise a strategy to reduce the number of required dichotomies in the overall multi-class process.

References

  1. Allwein, E., Shapire, R., and Singer, Y. (2000). Reducing multi-class to binary: A unifying approach for margin classifiers. JMLR, 1(1):113-141.
  2. Anand, R., Mehrotra, K., Mohan, C., and Ranka, S. (1995). Efficient classification for multi-class problems using modular neural networks. TNN, 6(1):117-124.
  3. Athisos, V., Stefan, A., Yuan, Q., and Sclaroff, S. (2007). Classmap: Efficient multiclass recognition via embeddings. In ICCV.
  4. Bishop, C. (2006). Pattern Recognition and Machine Learning. Springer, 1 edition.
  5. Clark, P. and Boswell, R. (1991). Rule induction with CN2: Some improvements. In EWSL, pages 151-163.
  6. Crammer, K. and Singer, Y. (2002). On the learnability and design of output codes for multi-class problems. JMLR, 47(2-3):201-233.
  7. Dietterich, T. and Bakiri, G. (1996). Solving multi-class problems via ECOC. JAIR, 2(1):263-286.
  8. Klautau, A., Jevtic, N., and Orlitsky, A. (2004). On nearestneighbor ECOC with application to all-pairs multiclass support vector machines. JMLR, 4(1):1-15.
  9. Moreira, M. and Mayoraz, E. (1998). Improved pairwise coupling classification with correcting classifiers. In ECML.
  10. Narasimhamrthy, A. (2005). Theoretical bounds of majority voting performance for a binary classification problem. TPAMI, 27(12):1988-1995.
  11. Passerini, A., Pontil, M., and Frasconi, P. (2004). New results on error correcting output codes of kernel machines. TNN, 15(1):45-54.
  12. Pedrajas, N. and Boyer, D. (2006). Improving multi-class pattern recognition by the combination of two strategies. TPAMI, 28(6):1001-1006.
  13. Platt, J., Christiani, N., and Taylor, J. (1999). Large margin dags for multi-class classification. In NIPS, pages 547-553.
  14. Pujol, O., Radeva, P., and Vitria, J. (2006). Discriminant ECOC: A heuristic method for application dependent design of ECOC. TPAMI, 28(6):1007-1012.
  15. Stehling, R., Nascimento, M., and Falca˜o, A. (2002). A compact and efficient image retrieval approach based on border/interior pixel classification. In CIKM, pages 102-109.
  16. Takenouchi, T. and Ishii, S. (2007). Multi-class classification as a decoding problem. In FOCI, pages 470-475.
  17. Windeatt, T. and Ghaderi, R. (2003). Coding and decoding strategies for multi-class learning problems. Information Fusion, 4(1):11-21.
  18. Young, C., Yen, C., Pao, Y., and Nagurka, M. (2006). Oneclass-at-time removal sequence planning method for multi-class problems. TNN, 17(6):1544-1549.
Download


Paper Citation


in Harvard Style

Rocha A. and Goldenstein S. (2009). MULTI-CLASS FROM BINARY - Divide to conquer . In Proceedings of the Fourth International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2009) ISBN 978-989-8111-69-2, pages 323-330. DOI: 10.5220/0001777803230330


in Bibtex Style

@conference{visapp09,
author={Anderson Rocha and Siome Goldenstein},
title={MULTI-CLASS FROM BINARY - Divide to conquer},
booktitle={Proceedings of the Fourth International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2009)},
year={2009},
pages={323-330},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001777803230330},
isbn={978-989-8111-69-2},
}


in EndNote Style

TY - CONF
JO - Proceedings of the Fourth International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, (VISIGRAPP 2009)
TI - MULTI-CLASS FROM BINARY - Divide to conquer
SN - 978-989-8111-69-2
AU - Rocha A.
AU - Goldenstein S.
PY - 2009
SP - 323
EP - 330
DO - 10.5220/0001777803230330