Abstract
One of the main factors affecting the effectiveness of ECOC methods for classification is the dependence among the errors of the computed codeword bits. We present an extensive experimental work for evaluating the dependence among output errors of the decomposition unit of ECOC learning machines. In particular, we compare the dependence between ECOC Multi Layer Perceptrons (ECOC monolithic), made up by a single MLP, and ECOC ensembles made up by a set of independent and parallel dichotomizers (ECOC PND), using measures based on mutual information. In this way we can analyze the relations between performances, design and dependence among output errors in ECOC learning machines. Results quantitatively show that the dependence among computed codeword bits is significantly smaller for ECOC PND, pointing out that ensembles of independent dichotomizers are better suited for implementing ECOC classification methods.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
E.L. Allwein, R.E. Schapire, and Y. Singer. Reducing multiclass to binary: a unifying approach for margin classifiers. In Proc. ICML’2000, The Seventeenth International Conference on Machine Learning, 2000.
A. Berger. Error correcting output coding for text classification. In IJCAI’99: Workshop on machine learning for information filtering, 1999.
Y. Crammer and Y. Singer. On the learnability and design of output codes for multiclass problems. In Proceedings of the Thirteenth Annual Conference on Computational Learning Theory, pages 35–46, 2000.
T.G. Dietterich and G. Bakiri. Solving multiclass learning problems via error-correcting output codes. Journal of Artificial Intelligence Research, (2):263–286, 1995.
R. Ghani. Using error correcting output codes for text classification. In ICML 2000: Proceedings of the 17th International Conference on Machine Learning, pages 303–310, San Francisco, US, 2000. Morgan Kaufmann Publishers.
V. Guruswami and A. Sahai. Multiclass learning, boosting, and error-correcting codes. In Proc. of the Twelfth Annual Conference on Computational Learning Theory, pages 145–155. ACM Press, 1999.
E. Kong and T.G. Dietterich. Error–correcting output coding correct bias and variance. In The XII International Conference on Machine Learning, pages 313–321, San Francisco, CA, 1995. Morgan Kauffman.
F. Masulli and G. Valentini. Effectiveness of error correcting output codes in multiclass learning problems. In Lecture Notes in Computer Science, volume 1857, pages 107–116. Springer-Verlag, Berlin, Heidelberg, 2000.
F. Masulli and G. Valentini. Mutual information methods for evaluating dependence among outputs in learning machines. Technical Report TR-01-02, DISI–Dipartimento di Informatica e Scienze dell’ Informazione–Università di Genova, 2001. ftp://ftp.disi.unige.it/person/ValentiniG/papers/TR-01-02.ps.gz.
E. Mayoraz and M. Moreira. On the decomposition of polychotomies into dichotomies. In The XIV International Conference on Machine Learning, pages 219–226, Nashville, TN, July 1997.
C.J. Merz and P.M. Murphy. UCI repository of machine learning databases, 1998. http://www.ics.uci.edu/mlearn/MLRepository.html.
W.W. Peterson and E.J. Jr.Weldon. Error correcting codes. MIT Press, Cambridge, MA, 1972.
G. Valentini and F. Masulli. NEURObjects, a set of library classes for neural networks development. In Proceedings of IIA’99 and SOCO’99, pages 184–190, Millet, Canada, 1999. ICSC Academic Press.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2001 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Masulli, F., Valentini, G. (2001). Dependence among Codeword Bits Errors in ECOC Learning Machines: An Experimental Analysis. In: Kittler, J., Roli, F. (eds) Multiple Classifier Systems. MCS 2001. Lecture Notes in Computer Science, vol 2096. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-48219-9_16
Download citation
DOI: https://doi.org/10.1007/3-540-48219-9_16
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-42284-6
Online ISBN: 978-3-540-48219-2
eBook Packages: Springer Book Archive