An investigation into the effect of ensemble size and voting threshold on the accuracy of neural network ensembles

Robert Cox; David Clark; Alice Richardson

doi:10.1007/3-540-46695-9_23

An investigation into the effect of ensemble size and voting threshold on the accuracy of neural network ensembles

Robert Cox, David Clark, Alice Richardson

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

4 Citations (Scopus)

Abstract

If voting is used by an ensemble to classify data, some data points may not be classified, but a higher proportion of those which are classified are classified correctly. This trade off is affected by ensemble size and voting threshold. This paper investigates the effect of ensemble size on the proportions of decisions made and correct decisions. It does this for majority voting and consensus voting on ensembles of neural network classifiers constructed using bagging. It also models the relationships in order to estimate the asymptotic values as the ensemble size increases.

Original language	English
Title of host publication	Advanced Topics in Artificial Intelligence - 12th Australian Joint Conference on Artificial Intelligence, AI 1999, Proceedings
Editors	Norman Foo
Publisher	Springer Verlag
Pages	268-277
Number of pages	10
ISBN (Print)	3540668225, 9783540668220
DOIs	https://doi.org/10.1007/3-540-46695-9_23
Publication status	Published - 1999
Event	12th Australian Joint Conference on Artificial Intelligence, AI 1999 - Sydney, Australia Duration: 6 Dec 1999 → 10 Dec 1999

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	1747
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	12th Australian Joint Conference on Artificial Intelligence, AI 1999
Country/Territory	Australia
City	Sydney
Period	6/12/99 → 10/12/99

Access to Document

10.1007/3-540-46695-9_23

Cite this

Cox, R., Clark, D., & Richardson, A. (1999). An investigation into the effect of ensemble size and voting threshold on the accuracy of neural network ensembles. In N. Foo (Ed.), Advanced Topics in Artificial Intelligence - 12th Australian Joint Conference on Artificial Intelligence, AI 1999, Proceedings (pp. 268-277). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1747). Springer Verlag. https://doi.org/10.1007/3-540-46695-9_23

Cox, Robert ; Clark, David ; Richardson, Alice. / An investigation into the effect of ensemble size and voting threshold on the accuracy of neural network ensembles. Advanced Topics in Artificial Intelligence - 12th Australian Joint Conference on Artificial Intelligence, AI 1999, Proceedings. editor / Norman Foo. Springer Verlag, 1999. pp. 268-277 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{48dfb7fd311649a18cd2e4f049752863,

title = "An investigation into the effect of ensemble size and voting threshold on the accuracy of neural network ensembles",

abstract = "If voting is used by an ensemble to classify data, some data points may not be classified, but a higher proportion of those which are classified are classified correctly. This trade off is affected by ensemble size and voting threshold. This paper investigates the effect of ensemble size on the proportions of decisions made and correct decisions. It does this for majority voting and consensus voting on ensembles of neural network classifiers constructed using bagging. It also models the relationships in order to estimate the asymptotic values as the ensemble size increases.",

author = "Robert Cox and David Clark and Alice Richardson",

note = "Publisher Copyright: {\textcopyright} Springer-Verlag Berlin Heidelberg 1999.; 12th Australian Joint Conference on Artificial Intelligence, AI 1999 ; Conference date: 06-12-1999 Through 10-12-1999",

year = "1999",

doi = "10.1007/3-540-46695-9_23",

language = "English",

isbn = "3540668225",

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

publisher = "Springer Verlag",

pages = "268--277",

editor = "Norman Foo",

booktitle = "Advanced Topics in Artificial Intelligence - 12th Australian Joint Conference on Artificial Intelligence, AI 1999, Proceedings",

address = "Germany",

}

Cox, R, Clark, D & Richardson, A 1999, An investigation into the effect of ensemble size and voting threshold on the accuracy of neural network ensembles. in N Foo (ed.), Advanced Topics in Artificial Intelligence - 12th Australian Joint Conference on Artificial Intelligence, AI 1999, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1747, Springer Verlag, pp. 268-277, 12th Australian Joint Conference on Artificial Intelligence, AI 1999, Sydney, Australia, 6/12/99. https://doi.org/10.1007/3-540-46695-9_23

An investigation into the effect of ensemble size and voting threshold on the accuracy of neural network ensembles. / Cox, Robert; Clark, David; Richardson, Alice.
Advanced Topics in Artificial Intelligence - 12th Australian Joint Conference on Artificial Intelligence, AI 1999, Proceedings. ed. / Norman Foo. Springer Verlag, 1999. p. 268-277 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1747).

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

TY - GEN

T1 - An investigation into the effect of ensemble size and voting threshold on the accuracy of neural network ensembles

AU - Cox, Robert

AU - Clark, David

AU - Richardson, Alice

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

PY - 1999

Y1 - 1999

N2 - If voting is used by an ensemble to classify data, some data points may not be classified, but a higher proportion of those which are classified are classified correctly. This trade off is affected by ensemble size and voting threshold. This paper investigates the effect of ensemble size on the proportions of decisions made and correct decisions. It does this for majority voting and consensus voting on ensembles of neural network classifiers constructed using bagging. It also models the relationships in order to estimate the asymptotic values as the ensemble size increases.

AB - If voting is used by an ensemble to classify data, some data points may not be classified, but a higher proportion of those which are classified are classified correctly. This trade off is affected by ensemble size and voting threshold. This paper investigates the effect of ensemble size on the proportions of decisions made and correct decisions. It does this for majority voting and consensus voting on ensembles of neural network classifiers constructed using bagging. It also models the relationships in order to estimate the asymptotic values as the ensemble size increases.

UR - http://www.scopus.com/inward/record.url?scp=84957633680&partnerID=8YFLogxK

U2 - 10.1007/3-540-46695-9_23

DO - 10.1007/3-540-46695-9_23

M3 - Conference contribution

AN - SCOPUS:84957633680

SN - 3540668225

SN - 9783540668220

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

SP - 268

EP - 277

BT - Advanced Topics in Artificial Intelligence - 12th Australian Joint Conference on Artificial Intelligence, AI 1999, Proceedings

A2 - Foo, Norman

PB - Springer Verlag

T2 - 12th Australian Joint Conference on Artificial Intelligence, AI 1999

Y2 - 6 December 1999 through 10 December 1999

ER -

Cox R, Clark D, Richardson A. An investigation into the effect of ensemble size and voting threshold on the accuracy of neural network ensembles. In Foo N, editor, Advanced Topics in Artificial Intelligence - 12th Australian Joint Conference on Artificial Intelligence, AI 1999, Proceedings. Springer Verlag. 1999. p. 268-277. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/3-540-46695-9_23

An investigation into the effect of ensemble size and voting threshold on the accuracy of neural network ensembles

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this