Skip to main content

Exact Learning Composed Classes with a Small Number of Mistakes

  • Conference paper
Learning Theory (COLT 2006)

Part of the book series: Lecture Notes in Computer Science ((LNAI,volume 4005))

Included in the following conference series:

  • 2865 Accesses

Abstract

The Composition Lemma is one of the strongest tools for learning complex classes. It shows that if a class is learnable then composing the class with a class of polynomial number of concepts gives a learnable class. In this paper we extend the Composition Lemma as follows: we show that composing an attribute efficient learnable class with a learnable class with polynomial shatter coefficient gives a learnable class.

This result extends many results in the literature and gives polynomial learning algorithms for new classes.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Institutional subscriptions

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Similar content being viewed by others

References

  1. Angluin, D.: Queries and Concept Learning. Machine Learning 2, 319–342 (1988)

    Google Scholar 

  2. Auer, P., Kwek, S., Maass, W., Warmuth, M.K.: Learning of Depth Two Neural Networks with Constant Fan-in at the Hidden Nodes. Electronic Colloquium on Computational Complexity (ECCC) 7(55) (2000)

    Google Scholar 

  3. Blum, A.: Rank-r Decision Trees are a Subclass of r-Decision Lists. Inf. Process. Lett. 42(4), 183–185 (1992)

    Article  MathSciNet  MATH  Google Scholar 

  4. Bshouty, N.H.: A Subexponential Exact Learning Algorithm for DNF Using Equivalence Queries. Inf. Process. Lett. 59(1), 37–39 (1996)

    Article  MathSciNet  MATH  Google Scholar 

  5. Bshouty, N.H.: A new Composition Theorem for Learning Algorithms. In: Proceedings of the 30th annual ACM Symposium on Theory of Computing (STOC), pp. 583–589 (1998)

    Google Scholar 

  6. Ehrenfeucht, A., Haussler, D.: Learning Decision Trees from Random Examples. Inf. Comput. 82(3), 231–246 (1989)

    Article  MathSciNet  MATH  Google Scholar 

  7. Kearns, M., Li, M., Pitt, L., Valiant, L.: On the learnability of boolean formulae. In: In Proceeding of the 19th ACM Symposium on the Theory of Computing, pp. 285–294 (1987)

    Google Scholar 

  8. Klivans, A.R., Servedio, R.A.: Learning DNF in time \(2^{\tilde O( n^{1/3})}\). J. Comput. Syst. Sci. 68(2), 303–318 (2004)

    Article  MathSciNet  MATH  Google Scholar 

  9. Littlestone, N.: Learning when irrelevant attributes abound. A new linear-threshold algorithm. Machine Learning 2, 285–318 (1988)

    Google Scholar 

  10. Maass, W., Warmuth, M.K.: Efficienct Learning with Virtual Threshold Gates. Information and Computation 141, 66–83 (1998)

    Article  MathSciNet  MATH  Google Scholar 

  11. Pitt, L., Warmuth, M.K.: Prediction-preserving reducibility. Journal of Computer and System Science 41(3), 430–467 (1990)

    Article  MathSciNet  MATH  Google Scholar 

  12. Sauer, N.: On the dencity of families of sets. J. Combinatorial Theory, Ser. A 13, 145–147 (1972)

    Article  MathSciNet  MATH  Google Scholar 

  13. Tarui, J., Tsukiji, T.: Learning DNF by Approximating Inclusion-Exclusion Formulae. In: IEEE Conference on Computational Complexity, pp. 215–220 (1999)

    Google Scholar 

  14. Valiant, L.G.: A theory of the learnable. Communication of the ACM 27(11) (1984)

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Editor information

Editors and Affiliations

Rights and permissions

Reprints and permissions

Copyright information

© 2006 Springer-Verlag Berlin Heidelberg

About this paper

Cite this paper

Bshouty, N.H., Mazzawi, H. (2006). Exact Learning Composed Classes with a Small Number of Mistakes. In: Lugosi, G., Simon, H.U. (eds) Learning Theory. COLT 2006. Lecture Notes in Computer Science(), vol 4005. Springer, Berlin, Heidelberg. https://doi.org/10.1007/11776420_17

Download citation

  • DOI: https://doi.org/10.1007/11776420_17

  • Publisher Name: Springer, Berlin, Heidelberg

  • Print ISBN: 978-3-540-35294-5

  • Online ISBN: 978-3-540-35296-9

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics