Advertisement

Rule induction with CN2: Some recent improvements

  • Peter Clark
  • Robin Boswell
Part 3: Numeric And Statistical Approaches
Part of the Lecture Notes in Computer Science book series (LNCS, volume 482)

Abstract

The CN2 algorithm induces an ordered list of classification rules from examples using entropy as its search heuristic. In this short paper, we describe two improvements to this algorithm. Firstly, we present the use of the Laplacian error estimate as an alternative evaluation function and secondly, we show how unordered as well as ordered rules can be generated. We experimentally demonstrate significantly improved performances resulting from these changes, thus enhancing the usefulness of CN2 as an inductive tool. Comparisons with Quinlan's C4.5 are also made.

Keywords

learning rule induction CN2 Laplace noise 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Buntine, W. and Niblett, T. A further comparison of splitting rules for decision-tree induction. (Submitted to the Machine Learning Journal), 1990.Google Scholar
  2. Cestnik, B. Estimating probabilities: A crucial task in machine learning. In ECAI-90, 1990.Google Scholar
  3. Cestnik, B. and Bratko, I. Learning redundant rules in noisy domains. In Kodratoff, Y., editor, ECAI-88, pages 348–350, London, Pitman, 1988.Google Scholar
  4. Cestnik, B. and Bratko, I. On estimating probabilities in tree pruning. In Kodratoff, Y., editor, Proc. EWSL-91, 1991.Google Scholar
  5. Clark, P. and Niblett, T. Induction in noisy domains. In Bratko, I. and Lavrač, N., editors, Progress in Machine Learning (proceedings of the 2nd European Working Session on Learning), Sigma, Wilmslow, UK, 1987.Google Scholar
  6. Clark, P. and Niblett, T. The CN2 induction algorithm. Machine Learning Journal, 3(4):261–283, 1989.Google Scholar
  7. Gams, M. New measurements highlight the importance of redundant knowledge. In Morik, K., editor, EWSL-89, pages 71–79, London, Pitman, 1989.Google Scholar
  8. Gams, M., Bohanec, M., and Cestnik, B. A schema for using multiple knowledge. (submitted to IJCAI-91), 1991.Google Scholar
  9. Hayes-Michie, J. E., editor Pragmatica: Bulletin of the Inductive Programming Special Interest Group, volume 1. Turing Institute Press, Glasgow, UK, 1990.Google Scholar
  10. Leech, W. J. A rule-based process control method with feedback. Advances in Instrumentation, 41:169–175, 1986.Google Scholar
  11. Michalski, R., Mozetic, I., Hong, J., and Lavrac, N. The multi-purpose incremental learning system AQ15 and its testing application to three medical domains. In AAAI-86, volume 2, pages 1041–1045, Ca. Kaufmann, 1986.Google Scholar
  12. Niblett, T. Constructing decision trees in noisy domains. In Bratko, I. and Lavrač, N., editors, Progress in Machine Learning (proceedings of the 2nd European Working Session on Learning), pages 67–78. Sigma, Wilmslow, UK, 1987.Google Scholar
  13. Quinlan, J. R. Simplifying decision trees. Int. Journal of Man-Machine Studies, 27(3):221–234, 1987.Google Scholar
  14. Quinlan, J. R., Compton, P. J., Horn, K. A., and Lazarus, L. Inductive knowledge acquisition: a case study. In Applications of Expert Systems, pages 157–173, Addison-Wesley, Wokingham, UK, 1987.Google Scholar
  15. Rivest, R. L. Learning decision lists. Machine Learning, 2(3):229–246, 1987.Google Scholar
  16. Slocombe, S., Moore, K., and Zelouf, M. Engineering expert system applications. In Annual Conference of the BCS Specialist Group on Expert Systems, 1986.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1991

Authors and Affiliations

  • Peter Clark
    • 1
  • Robin Boswell
    • 1
  1. 1.The Turing InstituteGlasgow

Personalised recommendations