Advertisement

Mitigating Concept Drift via Rejection

  • Jan Philip Göpfert
  • Barbara Hammer
  • Heiko Wersing
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11139)

Abstract

Learning in non-stationary environments is challenging, because under such conditions the common assumption of independent and identically distributed data does not hold; when concept drift is present it necessitates continuous system updates. In recent years, several powerful approaches have been proposed. However, these models typically classify any input, regardless of their confidence in the classification – a strategy, which is not optimal, particularly in safety-critical environments where alternatives to a (possibly unclear) decision exist, such as additional tests or a short delay of the decision. Formally speaking, this alternative corresponds to classification with rejection, a strategy which seems particularly promising in the context of concept drift, i.e. the occurrence of situations where the current model is wrong due to a concept change. In this contribution, we propose to extend learning under concept drift with rejection. Specifically, we extend two recent learning architectures for drift, the self-adjusting memory architecture (SAM-kNN) and adaptive random forests (ARF), to incorporate a reject option, resulting in highly competitive state-of-the-art technologies. We evaluate their performance in learning scenarios with different types of drift.

Keywords

Rejection Reject option Learning in non-stationary environments Concept drift 

References

  1. 1.
    Cha, E., Dragan, A.D., Srinivasa, S.S.: Perceived robot capability. In: 24th IEEE International Symposium on Robot and Human Interactive Communication, RO-MAN 2015, Kobe, Japan, August 31–September 4 2015, pp. 541–548 (2015)Google Scholar
  2. 2.
    Desai, M., et al.: Impact of robot failures and feedback on real-time trust. In: HRI. IEEE/ACM, pp. 251–258 (2013)Google Scholar
  3. 3.
    Kwon, M., Huang, S.H., Dragan, A.D.: Expressing robot incapability. In: Proceedings of the 2018 ACM/IEEE International Conference on Human-Robot Interaction, HRI 2018, Chicago, IL, USA, 05–08 March 2018, pp. 87–95 (2018)Google Scholar
  4. 4.
    Chow, C.: On optimum recognition error and reject tradeoff. IEEE Trans. Inf. Theor. 16(1), 41–46 (2006). ISSN 0018–9448MathSciNetCrossRefGoogle Scholar
  5. 5.
    Herbei, R., Wegkamp, M.H.: Classification with reject option. Can. J. Stat. 34(4), 709–721 (2006)MathSciNetCrossRefGoogle Scholar
  6. 6.
    Bartlett, P.L., Wegkamp, M.H.: Classification with a reject option using a hinge loss. J. Mach. Learn. Res. 9, 1823–1840 (2008). ISSN 1532–4435MathSciNetzbMATHGoogle Scholar
  7. 7.
    Villmann, T., et al.: Self-adjusting reject options in prototype based classification. In: Merényi, E., Mendenhall, M.J., O’Driscoll, P. (eds.) Advances in Self-organizing Maps and Learning Vector Quantization. AISC, vol. 428, pp. 269–279. Springer, Cham (2016).  https://doi.org/10.1007/978-3-319-28518-4_24CrossRefGoogle Scholar
  8. 8.
    Fischer, L., Hammer, B., Wersing, H.: Optimal local rejection for classifiers. Neurocomputing 214, 445–457 (2016)CrossRefGoogle Scholar
  9. 9.
    Vovk, V., Gammerman, A., Shafer, G.: Algorithmic Learning in a Random World. Springer, New York (2005).  https://doi.org/10.1007/b106715. ISBN 0387001522CrossRefzbMATHGoogle Scholar
  10. 10.
    Ditzler, G.: Learning in nonstationary environments: a survey. IEEE Comput. Intell. Mag. 10(4), 12–25 (2015). ISSN 1556–603XMathSciNetCrossRefGoogle Scholar
  11. 11.
    Gomes, H.M.: A survey on ensemble learning for data stream classification. ACM Comput. Surv. 50(2), 23:1–23:36 (2017)CrossRefGoogle Scholar
  12. 12.
    Losing, V., Hammer, B., Wersing, H.: Tackling heterogeneous concept drift with the Self-Adjusting Memory (SAM). Knowl. Inf. Syst. 54(1), 171–201 (2018)CrossRefGoogle Scholar
  13. 13.
    Loeffel, P.-X., Bifet, A., Marsala, C., Detyniecki, M.: Droplet ensemble learning on drifting data streams. In: Adams, N., Tucker, A., Weston, D. (eds.) IDA 2017. LNCS, vol. 10584, pp. 210–222. Springer, Cham (2017).  https://doi.org/10.1007/978-3-319-68765-0_18CrossRefGoogle Scholar
  14. 14.
    Gomes, H.M., et al.: Adaptive random forests for evolving data stream classification. Mach. Learn. 106, 1469–1495 (2017)MathSciNetCrossRefGoogle Scholar
  15. 15.
    Loeffel, P.X., Marsala, C., Detyniecki, M.: Classification with a reject option under concept drift: the droplets algorithm. In: 2015 IEEE International Conference on Data Science and Advanced Analytics (DSAA), pp. 1–9, October 2015Google Scholar
  16. 16.
    Platt, J.C.: Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods. In: Advances in Large Margin Classifiers, pp. 61–74. MIT Press (1999)Google Scholar
  17. 17.
    Hellman, M.E.: The nearest neighbor classification rule with a reject option. IEEE Trans. Syst. Sci. Cybern. 6(3), 179–185 (1970). ISSN 0536–1567CrossRefGoogle Scholar
  18. 18.
    Denoeux, T.: A k-nearest neighbor classification rule based on Dempster-Shafer theory. IEEE Trans. Syst. Man Cybern. 25(5), 804–813 (1995)CrossRefGoogle Scholar
  19. 19.
    Delany, S.J., Cunningham, P., Doyle, D., Zamolotskikh, A.: Generating estimates of classification confidence for a case-based spam filter. In: Muñoz-Ávila, H., Ricci, F. (eds.) ICCBR 2005. LNCS (LNAI), vol. 3620, pp. 177–190. Springer, Heidelberg (2005).  https://doi.org/10.1007/11536406_16CrossRefGoogle Scholar
  20. 20.
    Breiman, L.: Random forests. Mach. Learn. 45(1), 5–32 (2001). ISSN 0885–6125CrossRefGoogle Scholar
  21. 21.
    Fernández-Delgado, M.: Do we need hundreds of classifiers to solve real world classification problems? J. Mach. Learn. Res. 15, 3133–3181 (2014)MathSciNetzbMATHGoogle Scholar
  22. 22.
    Niculescu-Mizil, A., Caruana, R.: Predicting good probabilities with supervised learning. In: Proceedings of the 22nd International Conference on Machine Learning, ICML 2005, pp. 625–632. ACM, Bonn (2005). ISBN 1-59593-180-5Google Scholar
  23. 23.
    Nadeem, M.S.A., Zucker., Hanczar, B.: Accuracy-rejection curves (ARCs) for comparing classification methods with a reject option. In: Džeroski, S., Guerts, P., Rousu, J. (eds.) Proceedings of the Third International Workshop on Machine Learning in Systems Biology, Proceedings of Machine Learning Research, vol. 8, pp. 65–81. PMLR, Ljubljana (May 2009)Google Scholar
  24. 24.
    Bifet, A.: MOA: massive online analysis. J. Mach. Learn. Res. 11, 1601–1604 (2010). ISSN 1532–4435Google Scholar
  25. 25.
    Timothy, L.H., Watkin, A.R., Biehl, M.: The statistical mechanics of learning a rule. Rev. Mod. Phys. 65, 499–556 (1993)MathSciNetCrossRefGoogle Scholar
  26. 26.
    Losing, V., Hammer, B., Wersing, H.: KNN classifier with self adjusting memory for heterogeneous concept drift. In: 2016 IEEE 16th International Conference on Data Mining (ICDM), pp. 291–300. IEEE, Barcelona (2016)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  1. 1.Bielefeld University, Research Institute for Cognition and RoboticsBielefeldGermany
  2. 2.Honda Research Institute Europe GmbHOffenbachGermany

Personalised recommendations