OC1: randomized induction of oblique decision trees

Sreerama Murthy, Simon Kasif, Steven L Salzberg, Richard Beigel

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper introduces OC1, a new algorithm for generating multivariate decision trees. Multivariate trees classify examples by testing linear combinations of the features at each non-leaf node of the tree. Each test is equivalent to a hyperplane at an oblique orientation to the axes. Because of the computational intractability of finding an optimal orientation for these hyperplanes, heuristic methods must be used to produce good trees. This paper explores a new method that combines deterministic and randomized procedures to search for a good tree. Experiments on several different real-world data sets demonstrate that the method consistently finds much smaller trees than comparable methods using univariate tests. In addition, the accuracy of the trees found with our method matches or exceeds the best results of other machine learning methods.

Original languageEnglish (US)
Title of host publicationProceedings of the National Conference on Artificial Intelligence
PublisherPubl by AAAI
Pages322-327
Number of pages6
ISBN (Print)0262510715
StatePublished - 1993
EventProceedings of the 11th National Conference on Artificial Intelligence - Washington, DC, USA
Duration: Jul 11 1993Jul 15 1993

Other

OtherProceedings of the 11th National Conference on Artificial Intelligence
CityWashington, DC, USA
Period7/11/937/15/93

Fingerprint

Heuristic methods
Trees (mathematics)
Decision trees
Learning systems
Testing
Experiments

ASJC Scopus subject areas

  • Software

Cite this

Murthy, S., Kasif, S., Salzberg, S. L., & Beigel, R. (1993). OC1: randomized induction of oblique decision trees. In Proceedings of the National Conference on Artificial Intelligence (pp. 322-327). Publ by AAAI.

OC1 : randomized induction of oblique decision trees. / Murthy, Sreerama; Kasif, Simon; Salzberg, Steven L; Beigel, Richard.

Proceedings of the National Conference on Artificial Intelligence. Publ by AAAI, 1993. p. 322-327.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Murthy, S, Kasif, S, Salzberg, SL & Beigel, R 1993, OC1: randomized induction of oblique decision trees. in Proceedings of the National Conference on Artificial Intelligence. Publ by AAAI, pp. 322-327, Proceedings of the 11th National Conference on Artificial Intelligence, Washington, DC, USA, 7/11/93.
Murthy S, Kasif S, Salzberg SL, Beigel R. OC1: randomized induction of oblique decision trees. In Proceedings of the National Conference on Artificial Intelligence. Publ by AAAI. 1993. p. 322-327
Murthy, Sreerama ; Kasif, Simon ; Salzberg, Steven L ; Beigel, Richard. / OC1 : randomized induction of oblique decision trees. Proceedings of the National Conference on Artificial Intelligence. Publ by AAAI, 1993. pp. 322-327
@inproceedings{06ff9cc9688b407d8bf0c70c0e6f27db,
title = "OC1: randomized induction of oblique decision trees",
abstract = "This paper introduces OC1, a new algorithm for generating multivariate decision trees. Multivariate trees classify examples by testing linear combinations of the features at each non-leaf node of the tree. Each test is equivalent to a hyperplane at an oblique orientation to the axes. Because of the computational intractability of finding an optimal orientation for these hyperplanes, heuristic methods must be used to produce good trees. This paper explores a new method that combines deterministic and randomized procedures to search for a good tree. Experiments on several different real-world data sets demonstrate that the method consistently finds much smaller trees than comparable methods using univariate tests. In addition, the accuracy of the trees found with our method matches or exceeds the best results of other machine learning methods.",
author = "Sreerama Murthy and Simon Kasif and Salzberg, {Steven L} and Richard Beigel",
year = "1993",
language = "English (US)",
isbn = "0262510715",
pages = "322--327",
booktitle = "Proceedings of the National Conference on Artificial Intelligence",
publisher = "Publ by AAAI",

}

TY - GEN

T1 - OC1

T2 - randomized induction of oblique decision trees

AU - Murthy, Sreerama

AU - Kasif, Simon

AU - Salzberg, Steven L

AU - Beigel, Richard

PY - 1993

Y1 - 1993

N2 - This paper introduces OC1, a new algorithm for generating multivariate decision trees. Multivariate trees classify examples by testing linear combinations of the features at each non-leaf node of the tree. Each test is equivalent to a hyperplane at an oblique orientation to the axes. Because of the computational intractability of finding an optimal orientation for these hyperplanes, heuristic methods must be used to produce good trees. This paper explores a new method that combines deterministic and randomized procedures to search for a good tree. Experiments on several different real-world data sets demonstrate that the method consistently finds much smaller trees than comparable methods using univariate tests. In addition, the accuracy of the trees found with our method matches or exceeds the best results of other machine learning methods.

AB - This paper introduces OC1, a new algorithm for generating multivariate decision trees. Multivariate trees classify examples by testing linear combinations of the features at each non-leaf node of the tree. Each test is equivalent to a hyperplane at an oblique orientation to the axes. Because of the computational intractability of finding an optimal orientation for these hyperplanes, heuristic methods must be used to produce good trees. This paper explores a new method that combines deterministic and randomized procedures to search for a good tree. Experiments on several different real-world data sets demonstrate that the method consistently finds much smaller trees than comparable methods using univariate tests. In addition, the accuracy of the trees found with our method matches or exceeds the best results of other machine learning methods.

UR - http://www.scopus.com/inward/record.url?scp=0027709263&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0027709263&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0027709263

SN - 0262510715

SP - 322

EP - 327

BT - Proceedings of the National Conference on Artificial Intelligence

PB - Publ by AAAI

ER -