Advertisement

AD+Tree: A Compact Adaptation of Dynamic AD-Trees for Efficient Machine Learning on Large Data Sets

  • Jorge Moraleda
  • Teresa Miller
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2690)

Abstract

This paper introduces the AD+tree, a data structure for quickly counting the number of records that match conjunctive queries in a data set. The structure is useful for machine learning on large data sets. The AD+tree is an adaptation of the Dynamic AD-tree data structure [1].

We analyze the performance of AD+trees, comparing them to static AD-trees and Dynamic AD-trees. We show AD+trees maintain a very compact cache that enables them to handle queries on massively large data sets very efficiently even under complex, unstructured query patterns.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Komarek, P., Moore, A.W.: A dynamic adaptation of ad-trees for efficient machine learning on large data sets. In: Proceedings of the 17th International Conference on Machine Learning (2000)Google Scholar
  2. 2.
    Moore, A.W., Lee, M.S.: Cached sufficient statistics for efficient machine learning with large datasets. Journal of Artificial Intelligence Research 8 (1998)Google Scholar
  3. 3.
    Anderson, B., Moore, A.W.: Rich probabilistic models for gene expression. In: Proceedings of the 4th International Conference on Knowledge Discovery and Data Mining, pp. 134–138. AAAI Press, Menlo Park (1998)Google Scholar
  4. 4.
    Heckerman, D.: A tutorial for learning in bayesian networks, pp. 301–354. Kluwer Academic Publishers, Dordrecht (1998)Google Scholar
  5. 5.
    Dzeroski, S., Lavrac, N. (eds.): Relational Data Mining. Springer, Heidelberg (2001)zbMATHGoogle Scholar
  6. 6.
    Gaede, V., Günther, O.: Multidimensional access methods. ACM Computing Surveys 30(2), 170–231 (1998)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Jorge Moraleda
    • 1
  • Teresa Miller
    • 1
  1. 1.Aerospace Robotics LabStanford UniversityStanford

Personalised recommendations