Home > Research > Publications & Outputs > Federated $\mathcal{X}$-armed Bandit with Flexi...

Research

Associated organisational unit

Electronic data

  • 2409.07251v1

    Submitted manuscript, 562 KB, PDF document

    Available under license: CC BY: Creative Commons Attribution 4.0 International License

Links

Keywords

View graph of relations

Federated $\mathcal{X}$-armed Bandit with Flexible Personalisation

Research output: Working paper › Preprint

Published

Standard

Federated $\mathcal{X}$-armed Bandit with Flexible Personalisation. / Arabzadeh, Ali; Grant, James A.; Leslie, David S.
Arxiv, 2024.

Research output: Working paper › Preprint

Harvard

Arabzadeh, A, Grant, JA & Leslie, DS 2024 'Federated $\mathcal{X}$-armed Bandit with Flexible Personalisation' Arxiv. <https://arxiv.org/abs/2409.07251v1>

APA

Arabzadeh, A., Grant, J. A., & Leslie, D. S. (2024). Federated $\mathcal{X}$-armed Bandit with Flexible Personalisation. Arxiv. https://arxiv.org/abs/2409.07251v1

Vancouver

Arabzadeh A, Grant JA, Leslie DS. Federated $\mathcal{X}$-armed Bandit with Flexible Personalisation. Arxiv. 2024 Sept 11.

Author

Arabzadeh, Ali ; Grant, James A. ; Leslie, David S. / Federated $\mathcal{X}$-armed Bandit with Flexible Personalisation. Arxiv, 2024.

Bibtex

@techreport{2f6980737138448080e005efba9d8fab,
title = "Federated $\mathcal{X}$-armed Bandit with Flexible Personalisation",
abstract = "This paper introduces a novel approach to personalised federated learning within the $\mathcal{X}$-armed bandit framework, addressing the challenge of optimising both local and global objectives in a highly heterogeneous environment. Our method employs a surrogate objective function that combines individual client preferences with aggregated global knowledge, allowing for a flexible trade-off between personalisation and collective learning. We propose a phase-based elimination algorithm that achieves sublinear regret with logarithmic communication overhead, making it well-suited for federated settings. Theoretical analysis and empirical evaluations demonstrate the effectiveness of our approach compared to existing methods. Potential applications of this work span various domains, including healthcare, smart home devices, and e-commerce, where balancing personalisation with global insights is crucial.",
keywords = "stat.ML, cs.LG",
author = "Ali Arabzadeh and Grant, {James A.} and Leslie, {David S.}",
year = "2024",
month = sep,
day = "11",
language = "English",
publisher = "Arxiv",
type = "WorkingPaper",
institution = "Arxiv",

}

RIS

TY - UNPB

T1 - Federated $\mathcal{X}$-armed Bandit with Flexible Personalisation

AU - Arabzadeh, Ali

AU - Grant, James A.

AU - Leslie, David S.

PY - 2024/9/11

Y1 - 2024/9/11

N2 - This paper introduces a novel approach to personalised federated learning within the $\mathcal{X}$-armed bandit framework, addressing the challenge of optimising both local and global objectives in a highly heterogeneous environment. Our method employs a surrogate objective function that combines individual client preferences with aggregated global knowledge, allowing for a flexible trade-off between personalisation and collective learning. We propose a phase-based elimination algorithm that achieves sublinear regret with logarithmic communication overhead, making it well-suited for federated settings. Theoretical analysis and empirical evaluations demonstrate the effectiveness of our approach compared to existing methods. Potential applications of this work span various domains, including healthcare, smart home devices, and e-commerce, where balancing personalisation with global insights is crucial.

AB - This paper introduces a novel approach to personalised federated learning within the $\mathcal{X}$-armed bandit framework, addressing the challenge of optimising both local and global objectives in a highly heterogeneous environment. Our method employs a surrogate objective function that combines individual client preferences with aggregated global knowledge, allowing for a flexible trade-off between personalisation and collective learning. We propose a phase-based elimination algorithm that achieves sublinear regret with logarithmic communication overhead, making it well-suited for federated settings. Theoretical analysis and empirical evaluations demonstrate the effectiveness of our approach compared to existing methods. Potential applications of this work span various domains, including healthcare, smart home devices, and e-commerce, where balancing personalisation with global insights is crucial.

KW - stat.ML

KW - cs.LG

M3 - Preprint

BT - Federated $\mathcal{X}$-armed Bandit with Flexible Personalisation

PB - Arxiv

ER -