Optimal dyadic decision trees

G. Blanchard; C. Schäfer; Y. Rozenholc; K. R. Müller

doi:10.1007/s10994-007-0717-6

Optimal dyadic decision trees

G. Blanchard, C. Schäfer, Y. Rozenholc, K. R. Müller

Research output: Contribution to journal › Article › peer-review

24 Citations (Scopus)

Abstract

We introduce a new algorithm building an optimal dyadic decision tree (ODT). The method combines guaranteed performance in the learning theoretical sense and optimal search from the algorithmic point of view. Furthermore it inherits the explanatory power of tree approaches, while improving performance over classical approaches such as CART/C4.5, as shown on experiments on artificial and benchmark data.

Original language	English
Pages (from-to)	209-241
Number of pages	33
Journal	Machine Learning
Volume	66
Issue number	2-3
DOIs	https://doi.org/10.1007/s10994-007-0717-6
Publication status	Published - 2007 Mar

Keywords

Adaptive convergence rate
Classification
Decision tree
Density estimation
Oracle inequality

ASJC Scopus subject areas

Software
Artificial Intelligence

Access to Document

10.1007/s10994-007-0717-6

Fingerprint Dive into the research topics of 'Optimal dyadic decision trees'. Together they form a unique fingerprint.

Cite this

@article{5e5b3951f2854ec788f81dd55b7e07f2,

title = "Optimal dyadic decision trees",

abstract = "We introduce a new algorithm building an optimal dyadic decision tree (ODT). The method combines guaranteed performance in the learning theoretical sense and optimal search from the algorithmic point of view. Furthermore it inherits the explanatory power of tree approaches, while improving performance over classical approaches such as CART/C4.5, as shown on experiments on artificial and benchmark data.",

keywords = "Adaptive convergence rate, Classification, Decision tree, Density estimation, Oracle inequality",

author = "G. Blanchard and C. Sch{\"a}fer and Y. Rozenholc and M{\"u}ller, {K. R.}",

note = "Funding Information: Acknowledgments This work is partly funded by an grant of the Alexander von Humboldt Foundation, the PASCAL Network of Excellence (EU # 506778), and the Bundesministerium f{\"u}r Bildung und Forschung FKZ 01—BB02A and FKZ 01-SC40A. The authors thank Mikio Braun for valuable discussions, Nicolas Heefl for helping us with automatic tree drawing, and Alexander Binder for running the experiments again in Section 4.2 for the revision of the paper.",

year = "2007",

month = mar,

doi = "10.1007/s10994-007-0717-6",

language = "English",

volume = "66",

pages = "209--241",

journal = "Machine Learning",

issn = "0885-6125",

publisher = "Springer Netherlands",

number = "2-3",

}

TY - JOUR

T1 - Optimal dyadic decision trees

AU - Blanchard, G.

AU - Schäfer, C.

AU - Rozenholc, Y.

AU - Müller, K. R.

N1 - Funding Information: Acknowledgments This work is partly funded by an grant of the Alexander von Humboldt Foundation, the PASCAL Network of Excellence (EU # 506778), and the Bundesministerium für Bildung und Forschung FKZ 01—BB02A and FKZ 01-SC40A. The authors thank Mikio Braun for valuable discussions, Nicolas Heefl for helping us with automatic tree drawing, and Alexander Binder for running the experiments again in Section 4.2 for the revision of the paper.

PY - 2007/3

Y1 - 2007/3

N2 - We introduce a new algorithm building an optimal dyadic decision tree (ODT). The method combines guaranteed performance in the learning theoretical sense and optimal search from the algorithmic point of view. Furthermore it inherits the explanatory power of tree approaches, while improving performance over classical approaches such as CART/C4.5, as shown on experiments on artificial and benchmark data.

AB - We introduce a new algorithm building an optimal dyadic decision tree (ODT). The method combines guaranteed performance in the learning theoretical sense and optimal search from the algorithmic point of view. Furthermore it inherits the explanatory power of tree approaches, while improving performance over classical approaches such as CART/C4.5, as shown on experiments on artificial and benchmark data.

KW - Adaptive convergence rate

KW - Classification

KW - Decision tree

KW - Density estimation

KW - Oracle inequality

UR - http://www.scopus.com/inward/record.url?scp=33847623752&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33847623752&partnerID=8YFLogxK

U2 - 10.1007/s10994-007-0717-6

DO - 10.1007/s10994-007-0717-6

M3 - Article

AN - SCOPUS:33847623752

VL - 66

SP - 209

EP - 241

JO - Machine Learning

JF - Machine Learning

SN - 0885-6125

IS - 2-3

ER -