Empirical Comparison of Competing Query Learning Methods

  • Naoki Abe
  • Hiroshi Mamitsuka
  • Atsuyoshi Nakamura
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1532)

Abstract

Query learning is a form of machine learning in which the learner has control over the learning data it receives. In the context of discovery science, query learning may prove to be relevant in at least two ways. One is as a method of selective sampling, when a huge set of unlabeled data is available but a relatively small number of these data can be labeled, and a method that can selectively ask valuable queries is desired. The other is as a method of experimental design, where a query learning method is used to inform the experimenter what experiments are to be performed next

References

  1. 1.
    N. Abe, H. Li, and A. Nakamura. On-line learning of binary lexical relations using two-dimensional weighted majority algorithms. In Proc. 12th Int’l. Conference on Machine Learning, July 1995.Google Scholar
  2. 2.
    N. Abe and H. Mamitsuka. Query learning strategies using boosting and bagging. In Proceedings of the Fifteenth International Conference on Machine Learning, 1998.Google Scholar
  3. 3.
    D. Angluin. Learning regular sets from queries and counterexamples. Inform. Comput., 75(2):87–106, November 1987.Google Scholar
  4. 4.
    A. Nakamura and N. Abe. On-line learning of binary and n-ary relations over multidimensional clusters. In Proc. 8th Annu. Workshop on Comput. Learning Theory, July 1995.Google Scholar
  5. 5.
    G. Paass and J. Kindermann. Bayesian query construction for neural network models. In Advances in nueral information processing systems 7, pages 443–450, 1995.Google Scholar
  6. 6.
    H. S. Seung, M. Opper, and H. Sompolinsky. Query by committee. In Proc. 5th Annu. Workshop on Comput. Learning Theory, pages 287–294. ACM Press, New York, NY, 1992.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • Naoki Abe
    • 1
    • 2
  • Hiroshi Mamitsuka
    • 1
  • Atsuyoshi Nakamura
    • 1
  1. 1.NEC C& C Media Research LaboratoriesKawasakiJapan
  2. 2.Department of Computational Intelligence and Systems Sciences, Interdisciplinary Graduate School of Science and EngineeringTokyo Institute of TechnologyYokohamaJapan

Personalised recommendations