Cost-Sensitive Classification with Unconstrained Influence Diagrams

  • Jiří Iša
  • Zuzana Reitermanová
  • Ondřej Sýkora
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7147)

Abstract

In this paper, we deal with an enhanced problem of cost-sensitive classification, where not only the cost of misclassification needs to be minimized, but also the total cost of tests and their requirements. To solve this problem, we propose a novel method CS-UID based on the theory of Unconstrained Influence Diagrams (UIDs). We empirically evaluate and compare CS-UID with an existing algorithm for test-cost sensitive classification (TCSNB) on multiple real-world public referential datasets. We show that CS-UID outperforms TCSNB.

Keywords

Markov Decision Process Test Cost Decision Node Utility Variable Decision Tree Induction 
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.
This is a preview of subscription content, log in to check access

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Ahlmann-Ohlsen, K.S., Jensen, F.V., Nielsen, T.D., Pedersen, O., Vomlelová, M.: A Comparison of two Approaches for Solving Unconstrained Influence Diagrams. Int. J. of Approximate Reasoning 50(1), 153–173 (2009)MathSciNetCrossRefMATHGoogle Scholar
  2. 2.
    Chai, X., Deng, L., Yang, Q., Ling, C.X.: Test-Cost Sensitive Naive Bayes Classification. In: IEEE Int. Conf. on Data Mining, pp. 51–58 (2004)Google Scholar
  3. 3.
    Domingos, P.: MetaCost: A General Method for Making Classifiers Cost-Sensitive. In: Proc. 5th Int. Conf. on Knowledge Discovery and Data Mining, pp. 155–164 (1999)Google Scholar
  4. 4.
    Elkan, C.: The Foundations of Cost-Sensitive Learning. In: Proc. 17th Int. Joint Conf. on Artificial Intelligence, pp. 973–978 (2001)Google Scholar
  5. 5.
    Frank, A., Asuncion, A.: UCI machine learning repository (2010), http://archive.ics.uci.edu/ml
  6. 6.
    Fried, V.: Approximate solution of Unconstrained influence diagrams. Master’s thesis, Fac. of Math. and Phys., Charles Univ., Prague, Czech Republic (2006)Google Scholar
  7. 7.
    Greiner, R., Grove, A.J., Roth, D.: Learning Cost-sensitive Active Classifiers. Artificial Intelligence 139(2), 137–174 (2002)MathSciNetCrossRefGoogle Scholar
  8. 8.
    Iša, J., Lisý, V., Reitermanová, Z., Sýkora, O.: Unconstrained Influence Diagram Solver: Guido. In: Proc. 19th IEEE Int. Conf. on Tools with Artificial Intelligence, vol. 1, pp. 24–27. IEEE Computer Society, Washington, DC, USA (2007)Google Scholar
  9. 9.
    Iša, J., Reitermanová, Z., Sýkora, O.: On the Complexity of General Solution DAGs. In: Proc. 8th IEEE Int. Conf. on Machine Learning and Applications, pp. 673–678 (2009)Google Scholar
  10. 10.
    Jensen, F., Vomlelová, M.: Unconstrained Influence Diagrams. In: Proc. 18th Annu. Conf. on Uncertainty in Artificial Intelligence, pp. 234–241. Morgan Kaufmann, Edmonton (2002)Google Scholar
  11. 11.
    Jensen, F.V., Graven-Nielsen, T.: Bayesian Networks and Decision Graphs. Information Science and Statistics. Springer, New York (2007)CrossRefGoogle Scholar
  12. 12.
    Ling, C.X., Sheng, V.S., Yang, Q.: Test Strategies for Cost-Sensitive Decision Trees. IEEE Trans. on Knowledge Data Engineering 18(8), 1055–1067 (2006)CrossRefGoogle Scholar
  13. 13.
    Ling, C.X., Yang, Q., Wang, J., Zhang, S.: Decision Trees with Minimal Costs. In: Proc. 21st Int. Conf. on Machine Learning, pp. 4–8. Morgan Kaufmann (2004)Google Scholar
  14. 14.
    Luque, M., Nielsen, T.D., Jensen, F.V.: An Anytime Algorithm for Evaluating Unconstrained Influence Diagrams. In: Proc. 4th European Workshop on Probabilistic Graphical Models, Hirtshals, Denmark, pp. 177–184 (2008)Google Scholar
  15. 15.
    Núñez, M.: The Use of Background Knowledge in Decision Tree Induction. Machine Learning 6, 231–250 (1991)Google Scholar
  16. 16.
    Sheng, V.S., Ling, C.X., Ni, A., Zhang, S.: Cost-Sensitive Test Strategies. In: Proc. 21st Nat. Conf. on Artificial Intelligence, pp. 482–487. AAAI Press (2006)Google Scholar
  17. 17.
    Tan, M.: Cost-Sensitive Learning of Classification Knowledge and Its Applications in Robotics. Machine Learning 13, 7–33 (1993)Google Scholar
  18. 18.
    Ting, K.M.: Inducing Cost-Sensitive Trees via Instance Weighting. In: Proc. 2nd European Symposium Principles of Data Mining and Knowledge Discovery, pp. 139–147 (1998)Google Scholar
  19. 19.
    Ting, K.M.: An Instance-Weighting Method to Induce Cost-Sensitive Trees. IEEE Trans. on Knowledge Data Engineering 14(3), 659–665 (2002)CrossRefGoogle Scholar
  20. 20.
    Turney, P.D.: Cost-Sensitive Classification: Empirical Evaluation of a Hybrid Genetic Decision Tree Induction Algorithm. Journal of Artificial Intelligence Research, 369–409 (1995)Google Scholar
  21. 21.
    Zubek, V.B., Dietterich, T.G.: Integrating learning from examples into the search for diagnostic policies. Journal of Artificial Intelligence Research 24, 263–303 (2005)MATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Jiří Iša
    • 1
  • Zuzana Reitermanová
    • 1
  • Ondřej Sýkora
    • 1
  1. 1.Department of Theoretical Computer Science Faculty of Mathematics and PhysicsCharles University in PraguePragueCzech Republic

Personalised recommendations