Abstract
ECOC approach can be used to reduce a multiclass categorization problem to multiple binary problems and to improve the generalization of classifiers. Yet there is no single coding method that can generate ECOCs suitable for any number of classes. This paper provides a search-coding method that associates nonnegative integers with binary strings. Given any number of classes and an expected minimum hamming distance, the method can find out a satisfied output code through searching an integer range. Experimental results show that, as a general coding method, the search-coding method can improve the generalization for both stable and unstable classifiers efficiently
Preview
Unable to display preview. Download preview PDF.
Similar content being viewed by others
References
Rumelhart, D.E., Hinton, G.E., Williams, R.J.: Learning internal representations by error propagation. In: Parallel Distributed Processing-Explorations in the Microstructure of Cognition, ch. 8, pp. 318–362. MIT Press, Cambridge (1986)
Hastie, T., Tibshirani, R.: Classification by pairwise coupling. The Annals of Statistics 26.2, 451–471 (1998)
Sejnowski, T.J., Rosenberg, C.R.: Parallel networks that learn to pronounce english text. Journal of Complex Systems 1.1, 145–168 (1987)
Dietterich, T.G., Bakiri, G.: Solving multiclass learning problems via error-correcting output codes. Journal of Artificial Intelligence Research 3.2, 263–286 (1995)
Allwein, E., Schapire, R.E., Singer, Y.: Reducing multiclass to binary: A unifying approach for margin classifier. Journal of Machine Learning Research 1, 113–141 (2000)
Kong, E.B., Dietterich, T.G.: Error-Correcting Output Coding Corrects Bias and Variance. In: Proceedings of the twelfth International Conference on Machine Learning, California, USA, pp. 313–321 (1995)
Bay, S.D.: UCI KDD Archive (1999), http://kdd.ics.uci.edu
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2004 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Jiang, Yh., Zhao, Ql., Yang, Xj. (2004). A General Coding Method for Error-Correcting Output Codes. In: Dai, H., Srikant, R., Zhang, C. (eds) Advances in Knowledge Discovery and Data Mining. PAKDD 2004. Lecture Notes in Computer Science(), vol 3056. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-24775-3_76
Download citation
DOI: https://doi.org/10.1007/978-3-540-24775-3_76
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-22064-0
Online ISBN: 978-3-540-24775-3
eBook Packages: Springer Book Archive