Satisficing in Gaussian bandit problems

Paul Reverdy; Naomi E. Leonard

doi:10.1109/CDC.2014.7040284

Satisficing in Gaussian bandit problems

Paul Reverdy, Naomi E. Leonard

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

2 Scopus citations

Abstract

We propose a satisficing objective for the multi-armed bandit problem, i.e., where the objective is to achieve performance above a given threshold. We show that this new problem is equivalent to a standard multi-armed bandit problem with a maximizing objective and use this equivalence to find bounds on performance in terms of the satisficing objective. For the special case of Gaussian rewards we show that the satisficing problem is equivalent to a related standard multi-armed bandit problem again with Gaussian rewards. We apply the Upper Credible Limit (UCL) algorithm to this standard problem and show how it achieves optimal performance in terms of the satisficing objective.

Original language	English (US)
Title of host publication	53rd IEEE Conference on Decision and Control,CDC 2014
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	5718-5723
Number of pages	6
Edition	February
ISBN (Electronic)	9781479977468
DOIs	https://doi.org/10.1109/CDC.2014.7040284
State	Published - 2014
Externally published	Yes
Event	2014 53rd IEEE Annual Conference on Decision and Control, CDC 2014 - Los Angeles, United States Duration: Dec 15 2014 → Dec 17 2014

Publication series

Name	Proceedings of the IEEE Conference on Decision and Control
Number	February
Volume	2015-February
ISSN (Print)	0743-1546
ISSN (Electronic)	2576-2370

Conference

Conference	2014 53rd IEEE Annual Conference on Decision and Control, CDC 2014
Country/Territory	United States
City	Los Angeles
Period	12/15/14 → 12/17/14

ASJC Scopus subject areas

Control and Systems Engineering
Modeling and Simulation
Control and Optimization

Access to Document

10.1109/CDC.2014.7040284

Cite this

Reverdy, P., & Leonard, N. E. (2014). Satisficing in Gaussian bandit problems. In 53rd IEEE Conference on Decision and Control,CDC 2014 (February ed., pp. 5718-5723). Article 7040284 (Proceedings of the IEEE Conference on Decision and Control; Vol. 2015-February, No. February). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CDC.2014.7040284

Satisficing in Gaussian bandit problems. / Reverdy, Paul; Leonard, Naomi E.
53rd IEEE Conference on Decision and Control,CDC 2014. February. ed. Institute of Electrical and Electronics Engineers Inc., 2014. p. 5718-5723 7040284 (Proceedings of the IEEE Conference on Decision and Control; Vol. 2015-February, No. February).

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

Reverdy, P & Leonard, NE 2014, Satisficing in Gaussian bandit problems. in 53rd IEEE Conference on Decision and Control,CDC 2014. February edn, 7040284, Proceedings of the IEEE Conference on Decision and Control, no. February, vol. 2015-February, Institute of Electrical and Electronics Engineers Inc., pp. 5718-5723, 2014 53rd IEEE Annual Conference on Decision and Control, CDC 2014, Los Angeles, United States, 12/15/14. https://doi.org/10.1109/CDC.2014.7040284

@inproceedings{c40804476e8045029a746ab61f59e03f,

title = "Satisficing in Gaussian bandit problems",

abstract = "We propose a satisficing objective for the multi-armed bandit problem, i.e., where the objective is to achieve performance above a given threshold. We show that this new problem is equivalent to a standard multi-armed bandit problem with a maximizing objective and use this equivalence to find bounds on performance in terms of the satisficing objective. For the special case of Gaussian rewards we show that the satisficing problem is equivalent to a related standard multi-armed bandit problem again with Gaussian rewards. We apply the Upper Credible Limit (UCL) algorithm to this standard problem and show how it achieves optimal performance in terms of the satisficing objective.",

author = "Paul Reverdy and Leonard, {Naomi E.}",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 2014 53rd IEEE Annual Conference on Decision and Control, CDC 2014 ; Conference date: 15-12-2014 Through 17-12-2014",

year = "2014",

doi = "10.1109/CDC.2014.7040284",

language = "English (US)",

series = "Proceedings of the IEEE Conference on Decision and Control",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "February",

pages = "5718--5723",

booktitle = "53rd IEEE Conference on Decision and Control,CDC 2014",

edition = "February",

}

TY - GEN

T1 - Satisficing in Gaussian bandit problems

AU - Reverdy, Paul

AU - Leonard, Naomi E.

PY - 2014

Y1 - 2014

N2 - We propose a satisficing objective for the multi-armed bandit problem, i.e., where the objective is to achieve performance above a given threshold. We show that this new problem is equivalent to a standard multi-armed bandit problem with a maximizing objective and use this equivalence to find bounds on performance in terms of the satisficing objective. For the special case of Gaussian rewards we show that the satisficing problem is equivalent to a related standard multi-armed bandit problem again with Gaussian rewards. We apply the Upper Credible Limit (UCL) algorithm to this standard problem and show how it achieves optimal performance in terms of the satisficing objective.

AB - We propose a satisficing objective for the multi-armed bandit problem, i.e., where the objective is to achieve performance above a given threshold. We show that this new problem is equivalent to a standard multi-armed bandit problem with a maximizing objective and use this equivalence to find bounds on performance in terms of the satisficing objective. For the special case of Gaussian rewards we show that the satisficing problem is equivalent to a related standard multi-armed bandit problem again with Gaussian rewards. We apply the Upper Credible Limit (UCL) algorithm to this standard problem and show how it achieves optimal performance in terms of the satisficing objective.

UR - http://www.scopus.com/inward/record.url?scp=84988269471&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84988269471&partnerID=8YFLogxK

U2 - 10.1109/CDC.2014.7040284

DO - 10.1109/CDC.2014.7040284

M3 - Conference contribution

AN - SCOPUS:84988269471

T3 - Proceedings of the IEEE Conference on Decision and Control

SP - 5718

EP - 5723

BT - 53rd IEEE Conference on Decision and Control,CDC 2014

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2014 53rd IEEE Annual Conference on Decision and Control, CDC 2014

Y2 - 15 December 2014 through 17 December 2014

ER -

Satisficing in Gaussian bandit problems

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this