Advertisement

A New Way to Build Oblique Decision Trees Using Linear Programming

  • Guy Michel
  • Jean Luc Lambert
  • Bruno Cremilleux
  • Michel Henry-Amar
Part of the Studies in Classification, Data Analysis, and Knowledge Organization book series (STUDIES CLASS)

Summary

Adding linear combination splits to decision trees allows multivariate relations to be expressed more accurately and succinctly than univariate splits alone. In order to determine an oblique hyperplane which distinguishes two sets, linear programming is proposed to be used. This formulation yields a straightforward way to treat missing values. Computational comparison of that linear programming approach algorithm with classical univariate split algorithms proofs the interest of this method.

Key words

Oblique decision tree missing values linear programming 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    Bennet (1992). Decision tree construction via linear programming, Computer Sciences Technical report 1067.Google Scholar
  2. [2]
    Celeux (1988). Le traitement des valeurs manquantes dans le logiciel SICLA.Google Scholar
  3. [3]
    Chvatal (1993). Linear Programming, W.H. Freman and Compagny.Google Scholar
  4. [4]
    Mangasarian, Setiono & Wolberg (1990). Pattern recognition via linear programming: Theory and application to medical diagnosis, in: S.I.A.M. Workshop on optimisation.Google Scholar
  5. [5]
    Michie, Spiegelhalter & Taylor (1994). Machine learning, neural and statistical classification. Ellis Herwood Series in Artificial Intelligence.Google Scholar
  6. [6]
    Murthy, Kasif & Salzberg (1994). A system for induction of oblique decision trees in: Journal of Artificial Intelligence Research 2, 1–32.Google Scholar
  7. [7]
    Quinlan (1993). C4.5: Programs for machine learning. Morgan Kaufmann.Google Scholar
  8. [8]
    Quinlan (1989). Unknown attribute values in induction in Segre. Proceedings of the sixth International Workshop on Machine Learning, 164–168.Google Scholar

Copyright information

© Springer-Verlag Berlin · Heidelberg 1998

Authors and Affiliations

  • Guy Michel
    • 1
  • Jean Luc Lambert
    • 2
  • Bruno Cremilleux
    • 2
  • Michel Henry-Amar
    • 1
  1. 1.GREYCCNRS UPRESA 1772 Université de CaenCaen CedexFrance
  2. 2.CNRS UPRESA 6072 - GRECANCNRS UPRES 1772 Université de CaenCaen CedexFrance

Personalised recommendations