Asymptotic Optimality for Decentralised Bandits

Conor Newton; A J Ganesh; Henry Reeve

doi:10.1145/3512798.3512817

Asymptotic Optimality for Decentralised Bandits

Conor Newton, A J Ganesh, Henry Reeve

Research output: Contribution to journal › Article (Academic Journal) › peer-review

3 Citations (Scopus)

18 Downloads (Pure)

Abstract

We consider a large number of agents collaborating on a multi-armed bandit problem with a large number of arms. We present an algorithm which improves upon the Gossip- Insert-Eliminate method of Chawla et al. [3]. We provide a regret bound which shows that our algorithm is asymptotically optimal and present empirical results demonstrating lower regret on simulated data.

Original language	English
Pages (from-to)	51-53
Number of pages	3
Journal	ACM SIGMETRICS Performance Evaluation Review
Volume	49
Issue number	2
DOIs	https://doi.org/10.1145/3512798.3512817
Publication status	Published - 20 Jan 2022
Event	Reinforcement Learning in Networks and Queues, Sigmetrics 2021 - Duration: 14 Jun 2021 → 14 Jun 2021

Access to Document

10.1145/3512798.3512817Licence: Unspecified

Full-text PDF (accepted manuscript)
This is the accepted author manuscript (AAM). The final published version (Version of Record) can be found on the publisher's website. The copyright of any third-party content, such as images, remains with the copyright holder.
Accepted author manuscript, 352 KB

Handle.net

Persistent link

3 Citations
1 Article (Academic Journal)

Asymptotic Optimality for Decentralised Bandits
Newton, C. J., Ganesh, A. & Reeve, H. W. J., 1 Mar 2023, In: Dynamic Games and Applications. 13, p. 307-325
Research output: Contribution to journal › Article (Academic Journal) › peer-review

Open Access
File
1 Citation (Scopus)

104 Downloads (Pure)

Cite this

@article{13b2bfa1804543878c6a2572629a5217,

title = "Asymptotic Optimality for Decentralised Bandits",

abstract = "We consider a large number of agents collaborating on a multi-armed bandit problem with a large number of arms. We present an algorithm which improves upon the Gossip- Insert-Eliminate method of Chawla et al. [3]. We provide a regret bound which shows that our algorithm is asymptotically optimal and present empirical results demonstrating lower regret on simulated data. ",

author = "Conor Newton and Ganesh, \{A J\} and Henry Reeve",

year = "2022",

month = jan,

day = "20",

doi = "10.1145/3512798.3512817",

language = "English",

volume = "49",

pages = "51--53",

journal = "ACM SIGMETRICS Performance Evaluation Review",

issn = "0163-5999",

publisher = "Association for Computing Machinery",

number = "2",

note = "Reinforcement Learning in Networks and Queues, Sigmetrics 2021, RLNQ ; Conference date: 14-06-2021 Through 14-06-2021",

}

TY - JOUR

T1 - Asymptotic Optimality for Decentralised Bandits

AU - Newton, Conor

AU - Ganesh, A J

AU - Reeve, Henry

PY - 2022/1/20

Y1 - 2022/1/20

N2 - We consider a large number of agents collaborating on a multi-armed bandit problem with a large number of arms. We present an algorithm which improves upon the Gossip- Insert-Eliminate method of Chawla et al. [3]. We provide a regret bound which shows that our algorithm is asymptotically optimal and present empirical results demonstrating lower regret on simulated data.

AB - We consider a large number of agents collaborating on a multi-armed bandit problem with a large number of arms. We present an algorithm which improves upon the Gossip- Insert-Eliminate method of Chawla et al. [3]. We provide a regret bound which shows that our algorithm is asymptotically optimal and present empirical results demonstrating lower regret on simulated data.

U2 - 10.1145/3512798.3512817

DO - 10.1145/3512798.3512817

M3 - Article (Academic Journal)

SN - 0163-5999

VL - 49

SP - 51

EP - 53

JO - ACM SIGMETRICS Performance Evaluation Review

JF - ACM SIGMETRICS Performance Evaluation Review

IS - 2

T2 - Reinforcement Learning in Networks and Queues, Sigmetrics 2021

Y2 - 14 June 2021 through 14 June 2021

ER -

Asymptotic Optimality for Decentralised Bandits

Abstract

Access to Document

Handle.net

Fingerprint

Research output

Asymptotic Optimality for Decentralised Bandits

Cite this