The learnability of voting rules

Ariel D. Procaccia; Aviv Zohar; Yoni Peleg; Jeffrey S. Rosenschein

doi:10.1016/j.artint.2009.03.003

The learnability of voting rules

Ariel D. Procaccia^*, Aviv Zohar, Yoni Peleg, Jeffrey S. Rosenschein

^*Corresponding author for this work

The Rachel and Selim Benin School of Engineering and Computer Science

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

28 Scopus citations

Abstract

Scoring rules and voting trees are two broad and concisely-representable classes of voting rules; scoring rules award points to alternatives according to their position in the preferences of the voters, while voting trees are iterative procedures that select an alternative based on pairwise comparisons. In this paper, we investigate the PAC-learnability of these classes of rules. We demonstrate that the class of scoring rules, as functions from preferences into alternatives, is efficiently learnable in the PAC model. With respect to voting trees, while in general a learning algorithm would require an exponential number of samples, we show that if the number of leaves is polynomial in the size of the set of alternatives, then a polynomial training set suffices. We apply these results in an emerging theory: automated design of voting rules by learning.

Original language	English
Pages (from-to)	1133-1149
Number of pages	17
Journal	Artificial Intelligence
Volume	173
Issue number	12-13
DOIs	https://doi.org/10.1016/j.artint.2009.03.003
State	Published - Aug 2009

Bibliographical note

Funding Information:
The authors wish to thank anonymous AIJ, AAAI, and AAMAS reviewers, for their very helpful comments. This work was partially supported by Israel Science Foundation grant #898/05.

Funding Information:
✩ This paper subsumes two earlier conference papers [A.D. Procaccia, A. Zohar, Y. Peleg, J.S. Rosenschein, Learning voting trees, in: Proceedings of the 22nd AAAI Conference on AI (AAAI), 2007, pp. 110–115; A.D. Procaccia, A. Zohar, J.S. Rosenschein, Automated design of scoring rules by learning from examples, in: Proceedings of the 7th International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2008, pp. 951–958]. * Corresponding author. E-mail addresses: [email protected] (A.D. Procaccia), [email protected] (A. Zohar), [email protected] (Y. Peleg), [email protected] (J.S. Rosenschein). 1 The author was supported in this work by the Adams Fellowship Program of the Israel Academy of Sciences and Humanities.

Keywords

Computational learning theory
Computational social choice
Multiagent systems

Access to Document

10.1016/j.artint.2009.03.003

Cite this

@article{7a6f542e04244b1d82de6754628e6754,

title = "The learnability of voting rules",

abstract = "Scoring rules and voting trees are two broad and concisely-representable classes of voting rules; scoring rules award points to alternatives according to their position in the preferences of the voters, while voting trees are iterative procedures that select an alternative based on pairwise comparisons. In this paper, we investigate the PAC-learnability of these classes of rules. We demonstrate that the class of scoring rules, as functions from preferences into alternatives, is efficiently learnable in the PAC model. With respect to voting trees, while in general a learning algorithm would require an exponential number of samples, we show that if the number of leaves is polynomial in the size of the set of alternatives, then a polynomial training set suffices. We apply these results in an emerging theory: automated design of voting rules by learning.",

keywords = "Computational learning theory, Computational social choice, Multiagent systems",

author = "Procaccia, {Ariel D.} and Aviv Zohar and Yoni Peleg and Rosenschein, {Jeffrey S.}",

note = "Funding Information: The authors wish to thank anonymous AIJ, AAAI, and AAMAS reviewers, for their very helpful comments. This work was partially supported by Israel Science Foundation grant #898/05. Funding Information: ✩ This paper subsumes two earlier conference papers [A.D. Procaccia, A. Zohar, Y. Peleg, J.S. Rosenschein, Learning voting trees, in: Proceedings of the 22nd AAAI Conference on AI (AAAI), 2007, pp. 110–115; A.D. Procaccia, A. Zohar, J.S. Rosenschein, Automated design of scoring rules by learning from examples, in: Proceedings of the 7th International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2008, pp. 951–958]. * Corresponding author. E-mail addresses: [email protected] (A.D. Procaccia), [email protected] (A. Zohar), [email protected] (Y. Peleg), [email protected] (J.S. Rosenschein). 1 The author was supported in this work by the Adams Fellowship Program of the Israel Academy of Sciences and Humanities.",

year = "2009",

month = aug,

doi = "10.1016/j.artint.2009.03.003",

language = "אנגלית",

volume = "173",

pages = "1133--1149",

journal = "Artificial Intelligence",

issn = "0004-3702",

publisher = "Elsevier B.V.",

number = "12-13",

}

TY - JOUR

T1 - The learnability of voting rules

AU - Procaccia, Ariel D.

AU - Zohar, Aviv

AU - Peleg, Yoni

AU - Rosenschein, Jeffrey S.

N1 - Funding Information: The authors wish to thank anonymous AIJ, AAAI, and AAMAS reviewers, for their very helpful comments. This work was partially supported by Israel Science Foundation grant #898/05. Funding Information: ✩ This paper subsumes two earlier conference papers [A.D. Procaccia, A. Zohar, Y. Peleg, J.S. Rosenschein, Learning voting trees, in: Proceedings of the 22nd AAAI Conference on AI (AAAI), 2007, pp. 110–115; A.D. Procaccia, A. Zohar, J.S. Rosenschein, Automated design of scoring rules by learning from examples, in: Proceedings of the 7th International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2008, pp. 951–958]. * Corresponding author. E-mail addresses: [email protected] (A.D. Procaccia), [email protected] (A. Zohar), [email protected] (Y. Peleg), [email protected] (J.S. Rosenschein). 1 The author was supported in this work by the Adams Fellowship Program of the Israel Academy of Sciences and Humanities.

PY - 2009/8

Y1 - 2009/8

N2 - Scoring rules and voting trees are two broad and concisely-representable classes of voting rules; scoring rules award points to alternatives according to their position in the preferences of the voters, while voting trees are iterative procedures that select an alternative based on pairwise comparisons. In this paper, we investigate the PAC-learnability of these classes of rules. We demonstrate that the class of scoring rules, as functions from preferences into alternatives, is efficiently learnable in the PAC model. With respect to voting trees, while in general a learning algorithm would require an exponential number of samples, we show that if the number of leaves is polynomial in the size of the set of alternatives, then a polynomial training set suffices. We apply these results in an emerging theory: automated design of voting rules by learning.

AB - Scoring rules and voting trees are two broad and concisely-representable classes of voting rules; scoring rules award points to alternatives according to their position in the preferences of the voters, while voting trees are iterative procedures that select an alternative based on pairwise comparisons. In this paper, we investigate the PAC-learnability of these classes of rules. We demonstrate that the class of scoring rules, as functions from preferences into alternatives, is efficiently learnable in the PAC model. With respect to voting trees, while in general a learning algorithm would require an exponential number of samples, we show that if the number of leaves is polynomial in the size of the set of alternatives, then a polynomial training set suffices. We apply these results in an emerging theory: automated design of voting rules by learning.

KW - Computational learning theory

KW - Computational social choice

KW - Multiagent systems

UR - http://www.scopus.com/inward/record.url?scp=67449138583&partnerID=8YFLogxK

U2 - 10.1016/j.artint.2009.03.003

DO - 10.1016/j.artint.2009.03.003

M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???

AN - SCOPUS:67449138583

SN - 0004-3702

VL - 173

SP - 1133

EP - 1149

JO - Artificial Intelligence

JF - Artificial Intelligence

IS - 12-13

ER -

The learnability of voting rules

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this