On the complexity of samples for learning

Joel Ratsaby

doi:10.1007/978-3-540-27798-9_23

On the complexity of samples for learning

Joel Ratsaby

Research output: Chapter in Book/Report/Conference proceeding › Chapter › peer-review

Abstract

In machine-learning, maximizing the sample margin can reduce the learning generalization-error. Thus samples on which the target function has a large margin (γ) convey more information so we expect fewer such samples. In this paper, we estimate the complexity of a class of sets of large-margin samples for a general learning problem over a finite domain. We obtain an explicit dependence of this complexity on γ and the sample size.

Original language	English
Title of host publication	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Editors	Kyung-Yong Chwa, J. Ian Munro
Publisher	Springer Verlag
Pages	198-209
Number of pages	12
ISBN (Electronic)	354022856X, 9783540228561
DOIs	https://doi.org/10.1007/978-3-540-27798-9_23
State	Published - 1 Jan 2004
Externally published	Yes

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3106
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-540-27798-9_23

Cite this

Ratsaby, J. (2004). On the complexity of samples for learning. In K.-Y. Chwa, & J. Ian Munro (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 198-209). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3106). Springer Verlag. https://doi.org/10.1007/978-3-540-27798-9_23

Ratsaby, Joel. / On the complexity of samples for learning. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). editor / Kyung-Yong Chwa ; J. Ian Munro. Springer Verlag, 2004. pp. 198-209 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inbook{e956789096fe40b985314c152f8ce8e1,

title = "On the complexity of samples for learning",

abstract = "In machine-learning, maximizing the sample margin can reduce the learning generalization-error. Thus samples on which the target function has a large margin (γ) convey more information so we expect fewer such samples. In this paper, we estimate the complexity of a class of sets of large-margin samples for a general learning problem over a finite domain. We obtain an explicit dependence of this complexity on γ and the sample size.",

author = "Joel Ratsaby",

year = "2004",

month = jan,

day = "1",

doi = "10.1007/978-3-540-27798-9_23",

language = "English",

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

publisher = "Springer Verlag",

pages = "198--209",

editor = "Kyung-Yong Chwa and {Ian Munro}, J.",

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

address = "Germany",

}

Ratsaby, J 2004, On the complexity of samples for learning. in K-Y Chwa & J Ian Munro (eds), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3106, Springer Verlag, pp. 198-209. https://doi.org/10.1007/978-3-540-27798-9_23

On the complexity of samples for learning. / Ratsaby, Joel.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). ed. / Kyung-Yong Chwa; J. Ian Munro. Springer Verlag, 2004. p. 198-209 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3106).

Research output: Chapter in Book/Report/Conference proceeding › Chapter › peer-review

TY - CHAP

T1 - On the complexity of samples for learning

AU - Ratsaby, Joel

PY - 2004/1/1

Y1 - 2004/1/1

N2 - In machine-learning, maximizing the sample margin can reduce the learning generalization-error. Thus samples on which the target function has a large margin (γ) convey more information so we expect fewer such samples. In this paper, we estimate the complexity of a class of sets of large-margin samples for a general learning problem over a finite domain. We obtain an explicit dependence of this complexity on γ and the sample size.

AB - In machine-learning, maximizing the sample margin can reduce the learning generalization-error. Thus samples on which the target function has a large margin (γ) convey more information so we expect fewer such samples. In this paper, we estimate the complexity of a class of sets of large-margin samples for a general learning problem over a finite domain. We obtain an explicit dependence of this complexity on γ and the sample size.

UR - http://www.scopus.com/inward/record.url?scp=35048871696&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-27798-9_23

DO - 10.1007/978-3-540-27798-9_23

M3 - Chapter

AN - SCOPUS:35048871696

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

SP - 198

EP - 209

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

A2 - Chwa, Kyung-Yong

A2 - Ian Munro, J.

PB - Springer Verlag

ER -

Ratsaby J. On the complexity of samples for learning. In Chwa KY, Ian Munro J, editors, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Springer Verlag. 2004. p. 198-209. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-540-27798-9_23

On the complexity of samples for learning

Abstract

Publication series

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this