Building simple models: A case study with decision trees

David Jensen; Tim Oates; Paul R. Cohen

doi:10.1007/bfb0052842

Building simple models: A case study with decision trees

David Jensen, Tim Oates, Paul R. Cohen

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

3 Scopus citations

Abstract

Building correctly-sized models is a central challenge for induction algorithms. Many approaches to decision tree induction fail this challenge. Under a broad range of circumstances, these approaches exhibit a nearly linear relationship between training set size and tree size, even after accuracy has ceased to increase. These algorithms fail to adjust for the statistical effects of comparing multiple subtrees. Adjusting for these effects produces trees with little or no excess structure.

Original language	English (US)
Title of host publication	Advances in Intelligent Data Analysis
Subtitle of host publication	Reasoning about Data - 2nd International Symposium, IDA-1997, Proceedings
Editors	Xiaohui Liu, Paul Cohen, Michael Berthold
Publisher	Springer-Verlag
Pages	211-222
Number of pages	12
ISBN (Print)	9783540633464
DOIs	https://doi.org/10.1007/bfb0052842
State	Published - 1997
Externally published	Yes
Event	2nd International Symposium on Intelligent Data Analysis, IDA 1997 - London, United Kingdom Duration: Aug 4 1997 → Aug 6 1997

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	1280
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	2nd International Symposium on Intelligent Data Analysis, IDA 1997
Country/Territory	United Kingdom
City	London
Period	8/4/97 → 8/6/97

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/bfb0052842

Cite this

Jensen, D., Oates, T., & Cohen, P. R. (1997). Building simple models: A case study with decision trees. In X. Liu, P. Cohen, & M. Berthold (Eds.), Advances in Intelligent Data Analysis: Reasoning about Data - 2nd International Symposium, IDA-1997, Proceedings (pp. 211-222). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1280). Springer-Verlag. https://doi.org/10.1007/bfb0052842

Building simple models: A case study with decision trees. / Jensen, David; Oates, Tim; Cohen, Paul R.
Advances in Intelligent Data Analysis: Reasoning about Data - 2nd International Symposium, IDA-1997, Proceedings. ed. / Xiaohui Liu; Paul Cohen; Michael Berthold. Springer-Verlag, 1997. p. 211-222 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1280).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Jensen, D, Oates, T & Cohen, PR 1997, Building simple models: A case study with decision trees. in X Liu, P Cohen & M Berthold (eds), Advances in Intelligent Data Analysis: Reasoning about Data - 2nd International Symposium, IDA-1997, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1280, Springer-Verlag, pp. 211-222, 2nd International Symposium on Intelligent Data Analysis, IDA 1997, London, United Kingdom, 8/4/97. https://doi.org/10.1007/bfb0052842

Jensen D, Oates T, Cohen PR. Building simple models: A case study with decision trees. In Liu X, Cohen P, Berthold M, editors, Advances in Intelligent Data Analysis: Reasoning about Data - 2nd International Symposium, IDA-1997, Proceedings. Springer-Verlag. 1997. p. 211-222. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/bfb0052842

Jensen, David ; Oates, Tim ; Cohen, Paul R. / Building simple models : A case study with decision trees. Advances in Intelligent Data Analysis: Reasoning about Data - 2nd International Symposium, IDA-1997, Proceedings. editor / Xiaohui Liu ; Paul Cohen ; Michael Berthold. Springer-Verlag, 1997. pp. 211-222 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{441ea60fc5184e5baf83bdb54f7e1b32,

title = "Building simple models: A case study with decision trees",

abstract = "Building correctly-sized models is a central challenge for induction algorithms. Many approaches to decision tree induction fail this challenge. Under a broad range of circumstances, these approaches exhibit a nearly linear relationship between training set size and tree size, even after accuracy has ceased to increase. These algorithms fail to adjust for the statistical effects of comparing multiple subtrees. Adjusting for these effects produces trees with little or no excess structure.",

author = "David Jensen and Tim Oates and Cohen, {Paul R.}",

note = "Publisher Copyright: {\textcopyright} Springer-Verlag Berlin Heidelberg 1997.; 2nd International Symposium on Intelligent Data Analysis, IDA 1997 ; Conference date: 04-08-1997 Through 06-08-1997",

year = "1997",

doi = "10.1007/bfb0052842",

language = "English (US)",

isbn = "9783540633464",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer-Verlag",

pages = "211--222",

editor = "Xiaohui Liu and Paul Cohen and Michael Berthold",

booktitle = "Advances in Intelligent Data Analysis",

}

TY - GEN

T1 - Building simple models

T2 - 2nd International Symposium on Intelligent Data Analysis, IDA 1997

AU - Jensen, David

AU - Oates, Tim

AU - Cohen, Paul R.

N1 - Publisher Copyright: © Springer-Verlag Berlin Heidelberg 1997.

PY - 1997

Y1 - 1997

N2 - Building correctly-sized models is a central challenge for induction algorithms. Many approaches to decision tree induction fail this challenge. Under a broad range of circumstances, these approaches exhibit a nearly linear relationship between training set size and tree size, even after accuracy has ceased to increase. These algorithms fail to adjust for the statistical effects of comparing multiple subtrees. Adjusting for these effects produces trees with little or no excess structure.

AB - Building correctly-sized models is a central challenge for induction algorithms. Many approaches to decision tree induction fail this challenge. Under a broad range of circumstances, these approaches exhibit a nearly linear relationship between training set size and tree size, even after accuracy has ceased to increase. These algorithms fail to adjust for the statistical effects of comparing multiple subtrees. Adjusting for these effects produces trees with little or no excess structure.

UR - http://www.scopus.com/inward/record.url?scp=84867530100&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84867530100&partnerID=8YFLogxK

U2 - 10.1007/bfb0052842

DO - 10.1007/bfb0052842

M3 - Conference contribution

AN - SCOPUS:84867530100

SN - 9783540633464

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 211

EP - 222

BT - Advances in Intelligent Data Analysis

A2 - Liu, Xiaohui

A2 - Cohen, Paul

A2 - Berthold, Michael

PB - Springer-Verlag

Y2 - 4 August 1997 through 6 August 1997

ER -

Building simple models: A case study with decision trees

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this