Query by Committee, Linear Separation and Random Walks

Ran Bachrach; Shai Fine; Eli Shamir

doi:10.1007/3-540-49097-3_4

Query by Committee, Linear Separation and Random Walks

Ran Bachrach, Shai Fine, Eli Shamir

Hebrew University of Jerusalem

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

2 Scopus citations

Abstract

Recent works have shown the advantage of using Active Learning methods, such as the Query by Committee (QBC) algorithm, to various learning problems. This class of Algorithms requires an oracle with the ability to randomly select a consistent hypothesis according to some predefined distribution. When trying to implement such an oracle, for the linear separators family of hypotheses, various problems should be solved. The major problem is time-complexity, where the straight-forward Monte Carlo method takes exponential time.

Original language	English
Title of host publication	Computational Learning Theory
Editors	Paul Fischer, Hans Ulrich Simon
Place of Publication	Berlin, Heidelberg
Publisher	Springer Berlin Heidelberg
Pages	34-49
Number of pages	16
ISBN (Print)	978-3-540-49097-5
DOIs	https://doi.org/10.1007/3-540-49097-3_4
State	Published - 1999
Externally published	Yes
Event	4th European Conference, EuroCOLT’99 - Nordkirchen, Germany Duration: 29 Mar 1999 → 31 Mar 1999

Publication series

Name	Lecture Notes in Computer Science (LNCS), Lecture Notes in Artificial Intelligence (LNAI)
Publisher	Springer Berlin Heidelberg
Volume	1572
ISSN (Electronic)	0302-9743

Conference

Conference	4th European Conference, EuroCOLT’99
Country/Territory	Germany
Period	29/03/99 → 31/03/99

Access to Document

10.1007/3-540-49097-3_4

Cite this

@inproceedings{e9eb032b42cf4b96a204dd25d8b588bd,

title = "Query by Committee, Linear Separation and Random Walks",

abstract = "Recent works have shown the advantage of using Active Learning methods, such as the Query by Committee (QBC) algorithm, to various learning problems. This class of Algorithms requires an oracle with the ability to randomly select a consistent hypothesis according to some predefined distribution. When trying to implement such an oracle, for the linear separators family of hypotheses, various problems should be solved. The major problem is time-complexity, where the straight-forward Monte Carlo method takes exponential time.",

author = "Ran Bachrach and Shai Fine and Eli Shamir",

note = "Publisher Copyright: {\textcopyright} Springer-Verlag Berlin Heidelberg 1999.; 4th European Conference, EuroCOLT{\textquoteright}99 ; Conference date: 29-03-1999 Through 31-03-1999",

year = "1999",

doi = "10.1007/3-540-49097-3\_4",

language = "אנגלית",

isbn = "978-3-540-49097-5",

series = "Lecture Notes in Computer Science (LNCS), Lecture Notes in Artificial Intelligence (LNAI)",

publisher = "Springer Berlin Heidelberg",

pages = "34--49",

editor = "Paul Fischer and Simon, \{Hans Ulrich\}",

booktitle = "Computational Learning Theory",

}

Bachrach, R, Fine, S & Shamir, E 1999, Query by Committee, Linear Separation and Random Walks. in P Fischer & HU Simon (eds), Computational Learning Theory. Lecture Notes in Computer Science (LNCS), Lecture Notes in Artificial Intelligence (LNAI), vol. 1572, Springer Berlin Heidelberg, Berlin, Heidelberg, pp. 34-49, 4th European Conference, EuroCOLT’99, Germany, 29/03/99. https://doi.org/10.1007/3-540-49097-3_4

Query by Committee, Linear Separation and Random Walks. / Bachrach, Ran; Fine, Shai; Shamir, Eli.
Computational Learning Theory. ed. / Paul Fischer; Hans Ulrich Simon. Berlin, Heidelberg: Springer Berlin Heidelberg, 1999. p. 34-49 (Lecture Notes in Computer Science (LNCS), Lecture Notes in Artificial Intelligence (LNAI); Vol. 1572).

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

TY - GEN

T1 - Query by Committee, Linear Separation and Random Walks

AU - Bachrach, Ran

AU - Fine, Shai

AU - Shamir, Eli

N1 - Publisher Copyright: © Springer-Verlag Berlin Heidelberg 1999.

PY - 1999

Y1 - 1999

N2 - Recent works have shown the advantage of using Active Learning methods, such as the Query by Committee (QBC) algorithm, to various learning problems. This class of Algorithms requires an oracle with the ability to randomly select a consistent hypothesis according to some predefined distribution. When trying to implement such an oracle, for the linear separators family of hypotheses, various problems should be solved. The major problem is time-complexity, where the straight-forward Monte Carlo method takes exponential time.

AB - Recent works have shown the advantage of using Active Learning methods, such as the Query by Committee (QBC) algorithm, to various learning problems. This class of Algorithms requires an oracle with the ability to randomly select a consistent hypothesis according to some predefined distribution. When trying to implement such an oracle, for the linear separators family of hypotheses, various problems should be solved. The major problem is time-complexity, where the straight-forward Monte Carlo method takes exponential time.

UR - https://www.scopus.com/pages/publications/84947767124

U2 - 10.1007/3-540-49097-3_4

DO - 10.1007/3-540-49097-3_4

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

SN - 978-3-540-49097-5

T3 - Lecture Notes in Computer Science (LNCS), Lecture Notes in Artificial Intelligence (LNAI)

SP - 34

EP - 49

BT - Computational Learning Theory

A2 - Fischer, Paul

A2 - Simon, Hans Ulrich

PB - Springer Berlin Heidelberg

CY - Berlin, Heidelberg

T2 - 4th European Conference, EuroCOLT’99

Y2 - 29 March 1999 through 31 March 1999

ER -

Query by Committee, Linear Separation and Random Walks

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this