Optimal Extension of Error Correcting Output Codes

• Autores: Sergio Escalera, Oriol Pujol, Petia Radeva
• Localización: Artificial intelligence research and development / coord. por Beatriz López, Joaquim Meléndez Frigola , Petia Radeva Ivanova , Jordi Vitrià Marca , 2005, ISBN 978-1-58603-663-8, págs. 28-36
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • Error correcting output codes (ECOC) represent a successful extension of binary classifiers to address the multiclass problem. In this paper, we propose a novel technique called ECOC-ONE (Optimal Node Embedding) to improve an initial ECOC configuration defining a strategy to create new dichotomies and im-prove optimally the performance. The process of searching for new dichotomies is guided by the confusion matrices over two exclusive training subsets. A weighted methodology is proposed to take into account the different relevance between dichotomies. We validate our extension technique on well-known UCI databases. The results show significant improvement to the traditional coding techniques with far few extra cost.

• Referencias bibliográficas
  • [1] V. Vapnik, The nature of statistical learning theory, Springer-Verlag, 1995. Crossref Google Scholar
  • [2] J. Friedman, T. Hastie, R. Tibshirani, Additive logistic regression: a statistical view of boosting, The Annals of Statistics vol. 38...
  • [3] L. Breiman, J. Friedman, R. Olshen, C. Stone, Classification and Regression Trees, Wadsworth, 1984. Google Scholar
  • [4] E. Allwein, R. Schapire, Y. Singer, Reducing multiclass to binary: A unifying approach for margin classifiers, Journal of Machine Learning...
  • [5] E. Frank, S. Kramer, Ensembles of nested dichotomies for multiclass problems, in: Proceedings of 21st International Conference on Machine...
  • [6] T. Dietterich, G. Bakiri, Solving multiclass learning problems via error-correcting output codes, Journal of Artificial Intelligence Research...
  • [7] T. Windeatt, R. Ghaderi, Coding and decoding for multi-class learning problems, Information Fusion vol. 4 (1) (2003) 11-21. Crossref Google...
  • [8] T. Dietterich, E. Kong, Error-correcting output codes corrects bias and variance, in: P. of the 21th ICML (Ed.), S. Prieditis and S. Russell,...
  • [9] K. Crammer, Y. Singer, On the learnability and design of output codes for multiclass problems, Machine Learning 47 (2) (2002) 201-233. Digital...
  • [10] O. Pujol, P. Radeva, J. Vitrià, Discriminant (ecoc): A heuristic method for application dependent design of error correcting output codes,...
  • [11] T. Hastie, R. Tibshirani, Classification by pairwise grouping, The annals of statistics vol. 26 (5) (1998) 451-471. Google Scholar
  • [12] T. Dietterich, G. Bakiri, Error-correcting output codes: A general method for improving multiclass inductive learning programs, in: A....
  • [13] H. Madala, A. Ivakhnenko, Inductive Learning Algorithm for Complex Systems Modelling, CRC Press Inc, 1994. Digital Library Google Scholar
  • [14] A. Torralba, K. Murphy, W. Freeman, Sharing visual features for multiclass and multiview object detection, MIT AIM. Google Scholar