Advertisement

An Efficient Approximation to Lookahead in Relational Learners

  • Jan Struyf
  • Jesse Davis
  • David Page
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4212)

Abstract

Greedy machine learning algorithms suffer from shortsightedness, potentially returning suboptimal models due to limited exploration of the search space. Greedy search misses useful refinements that yield a significant gain only in conjunction with other conditions. Relational learners, such as inductive logic programming algorithms, are especially susceptible to this problem. Lookahead helps greedy search overcome myopia; unfortunately it causes an exponential increase in execution time. Furthermore, it may lead to overfitting. We propose a heuristic for greedy relational learning algorithms that can be seen as an efficient, limited form of lookahead. Our experimental evaluation shows that the proposed heuristic yields models that are as accurate as models generated using lookahead. It is also considerably faster than lookahead.

Keywords

Evaluation Score Inductive Logic Programming Relational Learner Greedy Search Account Number 
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.

References

  1. 1.
    Berka, P.: Guide to the financial data set. In: ECML/PKDD 2000 Discovery Challenge (2000)Google Scholar
  2. 2.
    Blockeel, H., De Raedt, L.: Top-down induction of first order logical decision trees. Artificial Intelligence 101(1-2), 285–297 (1998)MATHCrossRefMathSciNetGoogle Scholar
  3. 3.
    Blockeel, H., Dehaspe, L., Demoen, B., Janssens, G., Ramon, J., Vandecasteele, H.: Improving the efficiency of inductive logic programming through the use of query packs. Journal of Artificial Intelligence Research 16, 135–166 (2002)MATHGoogle Scholar
  4. 4.
    Blockeel, H., Struyf, J.: Frankenstein classifiers: Some experiments on the Sisyphus data set. In: Workshop on Integration of Data Mining, Decision Support, and Meta-Learning IDDM 2001 (2001)Google Scholar
  5. 5.
    Davis, J., Burnside, E., Dutra, I.C., Page, D., Santos Costa, V.: An integrated approach to learning bayesian networks of rules. In: 16th European Conference on Machine Learning, pp. 84–95 (2005)Google Scholar
  6. 6.
    Davis, J., Goadrich, M.: The relationship between precision-recall and ROC curves. In: 23nd International Conference on Machine Learning (to appear, 2006)Google Scholar
  7. 7.
    Džeroski, S., Lavrač, N. (eds.): Relational Data Mining. Springer, Heidelberg (2001)MATHGoogle Scholar
  8. 8.
    Mitchell, T.M.: Generalization as search. Artificial Intelligence 18, 203–226 (1982)CrossRefMathSciNetGoogle Scholar
  9. 9.
    Peña Castillo, L., Wrobel, S.: A comparative study on methods for reducing myopia of hill-climbing search in multirelational learning. In: 21st International Conference on Machine Learning (2004)Google Scholar
  10. 10.
    Quinlan, J.R.: C4.5: Programs for Machine Learning. Morgan Kaufmann series in Machine Learning. Morgan Kaufmann, San Francisco (1993)Google Scholar
  11. 11.
    Ray, S., Page, D.: Generalized skewing for functions with continuous and nominal attributes. In: 22nd International Conference on Machine Learning, pp. 705–712 (2005)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Jan Struyf
    • 1
  • Jesse Davis
    • 2
  • David Page
    • 2
  1. 1.Dept. of Computer ScienceKatholieke Universiteit LeuvenLeuvenBelgium
  2. 2.Dept. of Biostatistics and, Medical Informatics and Dept. of Computer SciencesUniversity of Wisconsin-MadisonMadisonUSA

Personalised recommendations