Federated Recommendation System via Differential Privacy

Tan Li; Linqi Song; Christina Fragouli

doi:10.1109/ISIT44484.2020.9174297

Federated Recommendation System via Differential Privacy

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

68 Citations (Scopus)

Abstract

In this paper we are interested in what we term the federated private bandits framework, that combines differential privacy with multi-agent bandit learning. We explore how differential privacy based Upper Confidence Bound (UCB) methods can be applied to multi-agent environments, and in particular to federated learning environments both in 'master-worker' and 'fully decentralized' settings. We provide theoretical analysis on the privacy and regret performance of the proposed methods and explore the tradeoffs between these two.

Original language	English
Title of host publication	2020 IEEE International Symposium on Information Theory - Proceedings
Publisher	IEEE
Pages	2592-2597
ISBN (Electronic)	978-1-7281-6432-8
DOIs	https://doi.org/10.1109/ISIT44484.2020.9174297
Publication status	Published - Jun 2020
Event	2020 IEEE International Symposium on Information Theory (ISIT 2020) - Virtual, Los Angeles, United States Duration: 21 Jun 2020 → 26 Jun 2020 https://2020.ieee-isit.org/

Publication series

Name	IEEE International Symposium on Information Theory - Proceedings
Volume	2020-June
ISSN (Print)	2157-8095

Conference

Conference	2020 IEEE International Symposium on Information Theory (ISIT 2020)
Abbreviated title	ISIT2020
Place	United States
City	Los Angeles
Period	21/06/20 → 26/06/20
Internet address	https://2020.ieee-isit.org/

Research Keywords

differential privacy
distributed learning
Federated learning
multi-arm bandit

Access Output

10.1109/ISIT44484.2020.9174297

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@inproceedings{4ff3f9890c2b497a95690753d1467842,

title = "Federated Recommendation System via Differential Privacy",

abstract = "In this paper we are interested in what we term the federated private bandits framework, that combines differential privacy with multi-agent bandit learning. We explore how differential privacy based Upper Confidence Bound (UCB) methods can be applied to multi-agent environments, and in particular to federated learning environments both in 'master-worker' and 'fully decentralized' settings. We provide theoretical analysis on the privacy and regret performance of the proposed methods and explore the tradeoffs between these two.",

keywords = "differential privacy, distributed learning, Federated learning, multi-arm bandit, differential privacy, distributed learning, Federated learning, multi-arm bandit, differential privacy, distributed learning, Federated learning, multi-arm bandit",

author = "Tan Li and Linqi Song and Christina Fragouli",

year = "2020",

month = jun,

doi = "10.1109/ISIT44484.2020.9174297",

language = "English",

series = "IEEE International Symposium on Information Theory - Proceedings",

publisher = "IEEE",

pages = "2592--2597",

booktitle = "2020 IEEE International Symposium on Information Theory - Proceedings",

address = "United States",

note = "2020 IEEE International Symposium on Information Theory (ISIT 2020), ISIT2020 ; Conference date: 21-06-2020 Through 26-06-2020",

url = "https://2020.ieee-isit.org/",

}

Li, T , Song, L & Fragouli, C 2020, Federated Recommendation System via Differential Privacy. in 2020 IEEE International Symposium on Information Theory - Proceedings., 9174297, IEEE International Symposium on Information Theory - Proceedings, vol. 2020-June, IEEE, pp. 2592-2597, 2020 IEEE International Symposium on Information Theory (ISIT 2020), Los Angeles, California, United States, 21/06/20. https://doi.org/10.1109/ISIT44484.2020.9174297

Federated Recommendation System via Differential Privacy. / Li, Tan ; Song, Linqi; Fragouli, Christina.
2020 IEEE International Symposium on Information Theory - Proceedings. IEEE, 2020. p. 2592-2597 9174297 (IEEE International Symposium on Information Theory - Proceedings; Vol. 2020-June).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - Federated Recommendation System via Differential Privacy

AU - Li, Tan

AU - Song, Linqi

AU - Fragouli, Christina

PY - 2020/6

Y1 - 2020/6

N2 - In this paper we are interested in what we term the federated private bandits framework, that combines differential privacy with multi-agent bandit learning. We explore how differential privacy based Upper Confidence Bound (UCB) methods can be applied to multi-agent environments, and in particular to federated learning environments both in 'master-worker' and 'fully decentralized' settings. We provide theoretical analysis on the privacy and regret performance of the proposed methods and explore the tradeoffs between these two.

AB - In this paper we are interested in what we term the federated private bandits framework, that combines differential privacy with multi-agent bandit learning. We explore how differential privacy based Upper Confidence Bound (UCB) methods can be applied to multi-agent environments, and in particular to federated learning environments both in 'master-worker' and 'fully decentralized' settings. We provide theoretical analysis on the privacy and regret performance of the proposed methods and explore the tradeoffs between these two.

KW - differential privacy

KW - distributed learning

KW - Federated learning

KW - multi-arm bandit

KW - differential privacy

KW - distributed learning

KW - Federated learning

KW - multi-arm bandit

KW - differential privacy

KW - distributed learning

KW - Federated learning

KW - multi-arm bandit

UR - http://www.scopus.com/inward/record.url?scp=85090413106&partnerID=8YFLogxK

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85090413106&origin=recordpage

U2 - 10.1109/ISIT44484.2020.9174297

DO - 10.1109/ISIT44484.2020.9174297

M3 - RGC 32 - Refereed conference paper (with host publication)

T3 - IEEE International Symposium on Information Theory - Proceedings

SP - 2592

EP - 2597

BT - 2020 IEEE International Symposium on Information Theory - Proceedings

PB - IEEE

T2 - 2020 IEEE International Symposium on Information Theory (ISIT 2020)

Y2 - 21 June 2020 through 26 June 2020

ER -

Federated Recommendation System via Differential Privacy

Abstract

Publication series

Conference

Research Keywords

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this