Advertisement

A Unifying Theory of Active Discovery and Learning

  • Timothy M. Hospedales
  • Shaogang Gong
  • Tao Xiang
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7576)

Abstract

For learning problems where human supervision is expensive, active query selection methods are often exploited to maximise the return of each supervision. Two problems where this has been successfully applied are active discovery – where the aim is to discover at least one instance of each rare class with few supervisions; and active learning – where the aim is to maximise a classifier’s performance with least supervision. Recently, there has been interest in optimising these tasks jointly, i.e., active learning with undiscovered classes, to support efficient interactive modelling of new domains. Mixtures of active discovery and learning and other schemes have been exploited, but perform poorly due to heuristic objectives. In this study, we show with systematic theoretical analysis how the previously disparate tasks of active discovery and learning can be cleanly unified into a single problem, and hence are able for the first time to develop a unified query algorithm to directly optimise this problem. The result is a model which consistently outperforms previous attempts at active learning in the presence of undiscovered classes, with no need to tune parameters for different datasets.

Keywords

Active Learning Active Discovery Unlabelled Data Minority Class Dirichlet Process 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Hospedales, T.M., Gong, S., Xiang, T.: Finding Rare Classes: Adapting Generative and Discriminative Models in Active Learning. In: Huang, J.Z., Cao, L., Srivastava, J. (eds.) PAKDD 2011, Part II. LNCS, vol. 6635, pp. 296–308. Springer, Heidelberg (2011)CrossRefGoogle Scholar
  2. 2.
    Haines, T., Xiang, T.: Active learning using dps for rare class discovery and classification. In: BMVC (2011)Google Scholar
  3. 3.
    Pelleg, D., Moore, A.: Active learning for anomaly and rare-category detection. In: NIPS (2004)Google Scholar
  4. 4.
    Stokes, J.W., Platt, J.C., Kravis, J., Shilman, M.: Aladin: Active learning of anomalies to detect intrusions. Technical Report 2008-24, MSR (2008)Google Scholar
  5. 5.
    Hospedales, T., Gong, S., Xiang, T.: Video behaviour mining using a dynamic topic model. International Journal of Computer Vision 98, 303–323 (2012)MathSciNetCrossRefGoogle Scholar
  6. 6.
    Guillaumin, M., Verbeek, J., Schmid, C.: Is that you? metric learning approaches for face identification. In: CVPR (2009)Google Scholar
  7. 7.
    Xiao, J., Hays, J., Ehinger, K.A., Oliva, A., Torralba, A.: Sun database: Large-scale scene recognition from abbey to zoo. In: CVPR, pp. 3485–3492 (2010)Google Scholar
  8. 8.
    Settles, B.: Active learning literature survey. Technical Report 1648, University of wisconsin–Madison (2009)Google Scholar
  9. 9.
    Tong, S., Koller, D.: Support vector machine active learning with applications to text classification. Journal of Machine Learning Research 2, 45–66 (2001)Google Scholar
  10. 10.
    Jain, P., Kapoor, A.: Active learning for large multi-class problems. In: CVPR (2009)Google Scholar
  11. 11.
    Cohn, D.A., Ghahramani, Z., Jordan, M.I.: Active learning with statistical models. Journal of Artificial Intelligence Research, 129–145 (1996)Google Scholar
  12. 12.
    Roy, N., McCallum, A.: Toward optimal active learning through sampling estimation of error reduction. In: ICML, pp. 441–448 (2001)Google Scholar
  13. 13.
    Kapoor, A., Horvitz, E., Basu, S.: Selective supervision: Guiding supervised learning with decision-theoretic active learning. In: IJCAI (2007)Google Scholar
  14. 14.
    Vijayanarasimhan, S., Grauman, K.: Multi-level active prediction of useful image annotations for recognition. In: NIPS (2008)Google Scholar
  15. 15.
    Beygelzimer, A., Hsu, D., Langford, J., Zhang, T.: Agnostic active learning without constraints. In: NIPS (2010)Google Scholar
  16. 16.
    He, J., Carbonell, J.: Nearest-neighbor-based active learning for rare category detection. In: NIPS (2007)Google Scholar
  17. 17.
    Vatturi, P., Wong, W.K.: Category detection using hierarchical mean shift. In: KDD, pp. 847–856 (2009)Google Scholar
  18. 18.
    Huang, H., He, Q., He, J., Ma, L.: RADAR: Rare Category Detection via Computation of Boundary Degree. In: Huang, J.Z., Cao, L., Srivastava, J. (eds.) PAKDD 2011, Part II. LNCS, vol. 6635, pp. 258–269. Springer, Heidelberg (2011)CrossRefGoogle Scholar
  19. 19.
    Antoniak, C.E.: Mixtures of dirichlet processes with applications to bayesian nonparametric problems. Annals of Statistics 2(6), 1152–1174 (1974)MathSciNetzbMATHCrossRefGoogle Scholar
  20. 20.
    Rasmussen, C.: The infinite gaussian mixture model. In: NIPS (2000)Google Scholar
  21. 21.
    Sillito, R., Fisher, R.: Incremental One-Class Learning with Bounded Computational Complexity. In: de Sá, J.M., Alexandre, L.A., Duch, W., Mandic, D.P. (eds.) ICANN 2007. LNCS, vol. 4668, pp. 58–67. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  22. 22.
    Escobar, M.D., West, M.: Bayesian density estimation and inference using mixtures. Journal of the American Statistical Association 90(430), 577–588 (1995)MathSciNetzbMATHCrossRefGoogle Scholar
  23. 23.
    Hospedales, T., Gong, S., Xiang, T.: Finding rare classes: Active learning with generative and discriminative models. In: IEEE TKDE (preprint)Google Scholar
  24. 24.
    Vijayanarasimhan, S., Grauman, K.: Cost-sensitive active visual category learning. International Journal of Computer Vision 91, 24–44 (2011)zbMATHCrossRefGoogle Scholar
  25. 25.
    Loy, C.C., Hospedales, T.M., Xiang, T., Gong, S.: Stream-based joint exploration-exploitation active learning. In: CVPR (2012)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Timothy M. Hospedales
    • 1
  • Shaogang Gong
    • 1
  • Tao Xiang
    • 1
  1. 1.EECSQueen Mary, University of LondonUK

Personalised recommendations