Skip to main content
SpringerLink
Log in

Polynomial Time Prediction Strategy with Almost Optimal Mistake Probability

  • Conference paper
Learning Theory (COLT 2004)

Part of the book series: Lecture Notes in Computer Science ((LNAI,volume 3120))

Included in the following conference series:

  • 2132 Accesses

Abstract

We give the first polynomial time prediction strategy for any PAC-learnable class C that probabilistically predicts the target with mistake probability

$$\frac{poly(log(t))}{t}=\widetilde{O}\frac{1}{t}$$

where t is the number of trials. The lower bound for the mistake probability is [HLW94] Ω(1/t) so our algorithm is almost optimal.

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

Access this chapter

Log in via an institution
Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 84.99
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 109.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Angluin, D.: Queries and concept learning. Machine Learning 2, 319–342 (1987)

    Google Scholar 

  2. Blum, A.: Separating distribution-free and mistake-bound learning models over the boolean domain. SIAM Journal on Computing 23(5), 990–1000 (1994)

    Article  MathSciNet  Google Scholar 

  3. Bshouty, N.H.: Exact learning of formulas in parallel. Machine Learning 26, 25–41 (1997)

    Article  MATH  Google Scholar 

  4. Bshouty, N.H., Cleve, R., Gavaldà, R., Kannan, S., Tamon, C.: Oracles and Queries That Are Sufficient for Exact Learning. Journal of Computer and System Sciences 52(3), 421–433 (1996)

    Article  MATH  MathSciNet  Google Scholar 

  5. Bshouty, N.H., Gavinsky, D.: PAC=PAExact and other equivalent models in learning. In: Proceedings of the 43rd Ann. Symp. on Foundation of Computer Science (FOCS), pp. 167–176 (2002)

    Google Scholar 

  6. Bshouty, N.H., Jackson, J., Tamon, C.: Exploring learnability between exact and PAC. In: Proceedings of the 15th Annual Conference on Computational Learning Theory, pp. 244–254 (2002)

    Google Scholar 

  7. Freund, Y.: Boosting a weak learning algorithm by majority. Information and Computation 121, 256–285 (1995)

    Article  MATH  MathSciNet  Google Scholar 

  8. Haussler, D., Littlestone, N., Warmuth, M.K.: Predicting 0,1-functions on randomly drawn points. Information and Computation 115, 248–292 (1994)

    Article  MATH  MathSciNet  Google Scholar 

  9. Kearns, M., Mansour, Y.: On the Boosting Ability of Top-Down Decision Tree Learning Algorithms. In: Proceedings of the 28th Symposium on Theory of Computing, pp. 459–468 (1996)

    Google Scholar 

  10. Littlestone, N.: Learning when irrelevant attributes abound: A new linearthreshold algorithm. Machine Learning 2, 285–318 (1988)

    Google Scholar 

  11. Mansour, Y., McAllester, D.: Boosting using Branching Programs. In: Proceedings of the 13th Annual Conference on Computational Learning Theory, pp. 220–224 (2000)

    Google Scholar 

  12. D. Gavinsky and A. Owshanko, PExact=Exact learning (manuscript)

    Google Scholar 

  13. Schapire, R.E.: The strength of weak learnability. Machine Learning 5(2), 197–227 (1990)

    Google Scholar 

  14. Valiant, L.: A theory of the learnable. Communications of the ACM 27(11), 1134–1142 (1984)

    Article  MATH  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Computer Science, Technion, Haifa, Israel, 32000

    Nader H. Bshouty

Authors
  1. Nader H. Bshouty
    View author publications

    You can also search for this author in PubMed Google Scholar

Editor information

Editors and Affiliations

  1. The Centre for Computational Statistics and Machine Learning Department of Computer Science, University College London, Gower St., WC1E 6BT, London

    John Shawe-Taylor

  2. Google, 1600 Amphitheater Parkway, CA 94043, Mountain View, USA

    Yoram Singer

Rights and permissions

Reprints and permissions

Copyright information

© 2004 Springer-Verlag Berlin Heidelberg

About this paper

Cite this paper

Bshouty, N.H. (2004). Polynomial Time Prediction Strategy with Almost Optimal Mistake Probability. In: Shawe-Taylor, J., Singer, Y. (eds) Learning Theory. COLT 2004. Lecture Notes in Computer Science(), vol 3120. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-27819-1_5

Download citation

  • DOI: https://doi.org/10.1007/978-3-540-27819-1_5

  • Publisher Name: Springer, Berlin, Heidelberg

  • Print ISBN: 978-3-540-22282-8

  • Online ISBN: 978-3-540-27819-1

  • eBook Packages: Springer Book Archive

Publish with us

Policies and ethics

Search

Navigation