Skip to main content

Classification on Soft Labels Is Robust against Label Noise

  • Conference paper
Knowledge-Based Intelligent Information and Engineering Systems (KES 2008)

Part of the book series: Lecture Notes in Computer Science ((LNAI,volume 5177))

Abstract

In a scenario of supervised classification of data, labeled training data is essential. Unfortunately, the process by which those labels are obtained is not error-free, for example due to human nature. The aim of this work is to find out what impact noise on the labels has, and we do so by artificially adding it. An algorithm for the noising procedure is described. Not only individual classifiers are studied, but also ensembles of classifiers whose answers are combined, increasing the overall performance. Also, we will answer the question if classifiers trained on soft labels are more resilient to label noise than those trained on hard labels.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Institutional subscriptions

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Similar content being viewed by others

References

  1. Steidl, S., Levit, M., Batliner, A., Nöth, E., Niemann, H.: ”Of all things the measure is man” - Automatic Classification of Emotions and Inter-Labeler Consistency. In: International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2005, pp. 317–320 (2005)

    Google Scholar 

  2. Thiel, C., Scherer, S., Schwenker, F.: Fuzzy-Input Fuzzy-Output One-Against-All Support Vector Machines. In: Apolloni, B., Howlett, R.J., Jain, L. (eds.) KES 2007, Part III. LNCS (LNAI), vol. 4694. Springer, Heidelberg (2007)

    Chapter  Google Scholar 

  3. Bondugula, R., Duzlevski, O., Xu, D.: Profiles and Fuzzy K-Nearest Neighbor Algorithm for Protein Secondary Structure Prediction. In: Chen, Y.P.P., Wong, L. (eds.) Proceedings of the 3rd Asia-Pacific Bioinformatics Conference, pp. 85–94. World Scientific, Singapore (2005)

    Chapter  Google Scholar 

  4. Angluin, D., Laird, P.: Learning from Noisy Examples. Machine Learning 2, 343–370 (1988)

    Google Scholar 

  5. Valiant, L.G.: A Theory of the Learnable. Commun. ACM 27, 1134–1142 (1984)

    Article  MATH  Google Scholar 

  6. Lawrence, N.D., Schölkopf, B.: Estimating a kernel Fisher discriminant in the presence of label noise. In: Proceedings of the 18th International Conference on Machine Learning, pp. 306–313. Morgan Kaufmann, San Francisco (2001)

    Google Scholar 

  7. Amini, M.R., Gallinari, P.: Semi-supervised learning with an explicit label-error model for misclassified data. In: IJCAI 2003 (2003)

    Google Scholar 

  8. McLachlan, G.J.: Discriminant Analysis and Statistical Pattern Recognition. John Wiley & Sons, Chichester (1992)

    Google Scholar 

  9. El Gayar, N., Schwenker, F., Palm, G.: A study of the robustness of KNN classifiers trained using soft labels. In: Schwenker, F., Marinai, S. (eds.) ANNPR 2006. LNCS (LNAI), vol. 4087, pp. 67–80. Springer, Heidelberg (2006)

    Chapter  Google Scholar 

  10. Fay, R., Kaufmann, U., Schwenker, F., Palm, G.: Learning Object Recognition in a NeuroBotic System. In: Groß, H.M., Debes, K., Böhme, H.J. (eds.) 3rd Workshop on SelfOrganization of AdaptiVE Behavior SOAVE 2004. Fortschritt-Berichte VDI, Reihe 10, vol. 743, pp. 198–209. VDI (2004)

    Google Scholar 

  11. Fay, R.: Feature Selection and Information Fusion in Hierarchical Neural Networks for Iterative 3D-Object Recognition. PhD thesis, University of Ulm, Germany (2007)

    Google Scholar 

  12. MacQueen, J.B.: Some methods for classification and analysis of multivariate observations. In: 5th Berkeley Symposium on Mathematical Statistics and Probability, pp. 281–298. University of California Press (1967)

    Google Scholar 

  13. Powell, M.J.D.: Radial basis functions for multivariate interpolation: A review. In: Mason, J.C., Cox, M.G. (eds.) Algorithms for Approximation, pp. 143–168. Clarendon Press, Oxford (1987)

    Google Scholar 

  14. Breiman, L., Meisel, W., Purcell, E.: Variable Kernel Estimates of Multivariate Densities. Technometrics 19, 135–144 (1977)

    Article  MATH  Google Scholar 

  15. Kahsay, L., Schwenker, F., Palm, G.: Comparison of multiclass SVM decomposition schemes for visual object recognition. In: Kropatsch, W.G., Sablatnig, R., Hanbury, A. (eds.) DAGM 2005. LNCS, vol. 3663, pp. 334–341. Springer, Heidelberg (2005)

    Chapter  Google Scholar 

  16. Kuncheva, L.I.: Combining Pattern Classifiers: Methods and Algorithms. Wiley, Chichester (2004)

    MATH  Google Scholar 

  17. Kuncheva, L.I., Bezdek, J.C., Duin, R.P.W.: Decision templates for multiple classifier fusion: An experimental comparison. Pattern Recognition 34, 299–314 (2001)

    Article  MATH  Google Scholar 

  18. Shafer, G.: Dempster-Shafer Theory (2002), http://www.glennshafer.com/assets/downloads/articles/article48.pdf

  19. Dempster, A.P.: A generalization of Bayesian inference. Journal of the Royal Statistical Society 30, 205–247 (1968)

    MathSciNet  Google Scholar 

  20. Shafer, G.: A Mathematical Theory of Evidence. University Press, Princeton (1976)

    MATH  Google Scholar 

  21. Schwenker, F., Dietrich, C., Thiel, C., Palm, G.: Learning decision fusion mappings for pattern recognition. ICGST International Journal on Artificial Intelligence and Machine Learning (AIML) 6, 17–21 (2006)

    Google Scholar 

  22. Binaghi, E., Brivio, P.A., Ghezzi, P., Rampini, A.: A fuzzy set-based accuracy assessment of soft classification. Pattern Recognition Letters 20, 935–948 (1999)

    Article  Google Scholar 

  23. Strauss, P.M., Hoffmann, H., Minker, W., Neumann, H., Palm, G., Scherer, S., Schwenker, F., Traue, H., Walter, W., Weidenbacher, U.: Wizard-of-oz data collection for perception and interaction in multi-user environments. In: International Conference on Language Resources and Evaluation (LREC) (2006)

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Editor information

Ignac Lovrek Robert J. Howlett Lakhmi C. Jain

Rights and permissions

Reprints and permissions

Copyright information

© 2008 Springer-Verlag Berlin Heidelberg

About this paper

Cite this paper

Thiel, C. (2008). Classification on Soft Labels Is Robust against Label Noise. In: Lovrek, I., Howlett, R.J., Jain, L.C. (eds) Knowledge-Based Intelligent Information and Engineering Systems. KES 2008. Lecture Notes in Computer Science(), vol 5177. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-85563-7_14

Download citation

  • DOI: https://doi.org/10.1007/978-3-540-85563-7_14

  • Publisher Name: Springer, Berlin, Heidelberg

  • Print ISBN: 978-3-540-85562-0

  • Online ISBN: 978-3-540-85563-7

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics