Reliability Maps: A Tool to Enhance Probability Estimates and Improve Classification Accuracy

  • Meelis Kull
  • Peter A. Flach
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8725)

Abstract

We propose a general method to assess the reliability of two-class probabilities in an instance-wise manner. This is relevant, for instance, for obtaining calibrated multi-class probabilities from two-class probability scores. The LS-ECOC method approaches this by performing least-squares fitting over a suitable error-correcting output code matrix, where the optimisation resolves potential conflicts in the input probabilities. While this gives all input probabilities equal weight, we would like to spend less effort fitting unreliable probability estimates. We introduce the concept of a reliability map to accompany the more conventional notion of calibration map; and LS-ECOC-R which modifies LS-ECOC to take reliability into account. We demonstrate on synthetic data that this gets us closer to the Bayes-optimal classifier, even if the base classifiers are linear and hence have high bias. Results on UCI data sets demonstrate that multi-class accuracy also improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Allwein, E., Schapire, R., Singer, Y.: Reducing multiclass to binary: A unifying approach for margin classifiers. Journal of Machine Learning Research 1, 113–141 (2001)MATHMathSciNetGoogle Scholar
  2. 2.
    Bella, A., Ferri, C., Hernández-Orallo, J., Ramírez-Quintana, M.J.: On the effect of calibration in classifier combination. Applied Intelligence 38(4), 566–585 (2012)CrossRefGoogle Scholar
  3. 3.
    Bennett, P.N.: Neighborhood-based local sensitivity. In: Kok, J.N., Koronacki, J., Lopez de Mantaras, R., Matwin, S., Mladenič, D., Skowron, A. (eds.) ECML 2007. LNCS (LNAI), vol. 4701, pp. 30–41. Springer, Heidelberg (2007)Google Scholar
  4. 4.
    Dietterich, T., Bakiri, G.: Solving Multiclass Learning Problems via Error-Correcting Output Codes. Journal of Artificial Intelligence Research 2, 263–286 (1995)MATHGoogle Scholar
  5. 5.
    Fan, J., Yao, Q.: Efficient estimation of conditional variance functions in stochastic regression. Biometrika 85(3), 645–660 (1998)CrossRefMATHMathSciNetGoogle Scholar
  6. 6.
    Galar, M., Fernández, A., Barrenechea, E., Bustince, H., Herrera, F.: Dynamic classifier selection for one-vs-one strategy. Pattern Recognition 46, 3412–3424 (2013)CrossRefGoogle Scholar
  7. 7.
    Kaufman, L., Rousseeuw, P.J.: Finding groups in data: an introduction to cluster analysis, vol. 344. John Wiley & Sons (2009)Google Scholar
  8. 8.
    Kong, E.B., Dietterich, T.: Probability estimation via error-correcting output coding. In: International Conference of Artificial Intelligence and Soft Computing, Banff, Canada (1997)Google Scholar
  9. 9.
    Lorena, A.C., Carvalho, A., Gama, J.: A review on the combination of binary classifiers in multiclass problems. Artificial Intelligence Review 30, 19–37 (2009)CrossRefGoogle Scholar
  10. 10.
    Murphy, A.H.: A new vector partition of the probability score. Journal of Applied Meteorology 12(4), 595–600 (1973)CrossRefGoogle Scholar
  11. 11.
    O’Brien, D., Gupta, M., Gray, R.: Cost-sensitive multi-class classification from probability estimates. In: Proceedings of the 25th International Conference on Machine Learning, Helsinki, Finland, pp. 712–719 (2008)Google Scholar
  12. 12.
    Schapire, R.E., Singer, Y.: Improved boosting algorithms using confidence-rated predictions. Machine Learning 37(3), 297–336 (1999)CrossRefMATHGoogle Scholar
  13. 13.
    Shafer, G., Vovk, V.: A tutorial on conformal prediction. Journal of Machine Learning Research 9, 371–421 (2008)MATHMathSciNetGoogle Scholar
  14. 14.
    Székely, G.J., Rizzo, M.L.: A new test for multivariate normality. Journal of Multivariate Analysis 93(1), 58–80 (2005)CrossRefMATHMathSciNetGoogle Scholar
  15. 15.
    Wang, X., Zhou, J.: Research on the characteristic of the probabilistic outputs via LS-ECOC. In: Eighth International Conference on Fuzzy Systems and Knowledge Discovery, vol. 2, pp. 1330–1334. IEEE (2011)Google Scholar
  16. 16.
    Windeatt, T., Ghaderi, R.: Coding and decoding strategies for multi-class learning problems. Information Fusion 4(1), 11–21 (2003)CrossRefGoogle Scholar
  17. 17.
    Zadrozny, B., Elkan, C.: Transforming classifier scores into accurate multiclass probability estimates. In: Eighth ACM SIGKDD International Conference, pp. 694–699. ACM Press, New York (2002)Google Scholar
  18. 18.
    Zhou, J.D., Wang, X.D., Song, H.: Research on the unbiased probability estimation of error-correcting output coding. Pattern Recognition 44(7), 1552–1565 (2011)CrossRefMATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2014

Authors and Affiliations

  • Meelis Kull
    • 1
  • Peter A. Flach
    • 1
  1. 1.Intelligent Systems LaboratoryUniversity of BristolUnited Kingdom

Personalised recommendations