Summary
The notion of a weak classifier, as one which is “a little better” than a random one, was introduced first for 2-class problems [1]. The extensions to K-class problems are known. All are based on relative activations for correct and incorrect classes and do not take into account the final choice of the answer. A new understanding and definition is proposed here. It takes into account only the final choice of classification that must be taken. It is shown that for a K class classifier to be called “weak”, it needs to achieve lower than 1/K risk value. This approach considers only the probability of the final answer choice, not the actual activations.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Kearns, M., Valiant, L.: Cryptographic limitations on learning Boolean formulae and finite automata. Journal of the Association for Computing Machinery 41(1), 67–95
Bax, E.: Validation of Voting Committees. Neural Computation 4(10), 975–986 (1998)
Hastie, T., Tibshirani, R., Friedman, J.: The Elements of Statistical Learning. Springer, New York (2001)
Tresp, V.: Committee Machines. In: Hu, Y.H., Hwang, J.-N. (eds.) Handbook for Neural Network Signal Processing. CRC Press, Boca Raton (2001)
Kittler, J., Hojjatoleslami, A., Windeatt, T.: Strategies for combining classifiers employing shared and distinct pattern representations. Pattern Recognition Letters 18, 1373–1377 (1997)
Schapire, R.E.: The strength of weak learnability. Machine Learning 5, 197–227 (1990)
Eibl, G., Pfeiffer, K.-P.: Multiclass boosting for weak classifiers. Journal of Machine Learning 6, 189–210 (2005)
Freund, Y., Schapire, R.E.: A decision theoretic generalization of online learning and an application to boosting. Journal of Computer and System Sciences 55, 119–139 (1997)
Podolak, I.T.: Hierarchical Classifier with Overlapping Class Groups. Expert Systems with Applications 34(1), 673–682 (2008)
Podolak, I.T., Biel, S.: Hierarchical classifier. In: Wyrzykowski, R., Dongarra, J., Meyer, N., Waśniewski, J. (eds.) PPAM 2005. LNCS, vol. 3911, pp. 591–598. Springer, Heidelberg (2006)
Podolak, I.T.: Hierarchical rules for a hierarchical classifier. In: Beliczynski, B., Dzielinski, A., Iwanowski, M., Ribeiro, B. (eds.) ICANNGA 2007. LNCS, vol. 4431, pp. 749–757. Springer, Heidelberg (2007)
Podolak, I.T., Roman, A.: Improving the accuracy of a hierarchical classifier (in preparation)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Podolak, I.T., Roman, A. (2009). A New Notion of Weakness in Classification Theory. In: Kurzynski, M., Wozniak, M. (eds) Computer Recognition Systems 3. Advances in Intelligent and Soft Computing, vol 57. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-93905-4_29
Download citation
DOI: https://doi.org/10.1007/978-3-540-93905-4_29
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-93904-7
Online ISBN: 978-3-540-93905-4
eBook Packages: EngineeringEngineering (R0)