BC-FL k-means: A Blockchain-based Framework for Federated Clustering

Mina Alishahi; Wouter Leeuw; Nicola Zannone

doi:10.1109/TrustCom60117.2023.00184

BC-FL k-means: A Blockchain-based Framework for Federated Clustering

Mina Alishahi, Wouter Leeuw, Nicola Zannone

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/Congresprocedure › Conferentiebijdrage › Academic › peer review

Samenvatting

This work presents a novel framework to train clustering models collaboratively without compromising accuracy while accommodating privacy and security in a decentralized manner. Our decentralized collaborative learning model removes the single point of failure and excludes unreliable input by designing a committee-based consensus method in a blockchain-based federated learning, which is equipped with a reputation system. We present a prototype implementation of our approach and show that its performance is comparable with centralized clustering regardless of the distribution of data among devices.

Originele taal-2	Engels
Titel	2023 IEEE 22nd International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom)
Redacteuren	Jia Hu, Geyong Min, Guojun Wang
Uitgeverij	Institute of Electrical and Electronics Engineers
Pagina's	1348-1354
Aantal pagina's	7
ISBN van elektronische versie	9798350381993
DOI's	https://doi.org/10.1109/TrustCom60117.2023.00184
Status	Gepubliceerd - 29 mei 2024
Evenement	22nd IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2023 - Exeter, Verenigd Koninkrijk Duur: 1 nov. 2023 → 3 nov. 2023

Congres

Congres	22nd IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2023
Land/Regio	Verenigd Koninkrijk
Stad	Exeter
Periode	1/11/23 → 3/11/23

Bibliografische nota

Publisher Copyright:
© 2023 IEEE.

Toegang tot document

10.1109/TrustCom60117.2023.00184

Andere bestanden en links

Link naar publicatie in Scopus

Citeer dit

Alishahi, M., Leeuw, W., & Zannone, N. (2024). BC-FL k-means: A Blockchain-based Framework for Federated Clustering. In J. Hu, G. Min, & G. Wang (editors), 2023 IEEE 22nd International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom) (blz. 1348-1354). Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/TrustCom60117.2023.00184

@inproceedings{1ad1874a74f9415680c9f7c87b1c248d,

title = "BC-FL k-means: A Blockchain-based Framework for Federated Clustering",

abstract = "This work presents a novel framework to train clustering models collaboratively without compromising accuracy while accommodating privacy and security in a decentralized manner. Our decentralized collaborative learning model removes the single point of failure and excludes unreliable input by designing a committee-based consensus method in a blockchain-based federated learning, which is equipped with a reputation system. We present a prototype implementation of our approach and show that its performance is comparable with centralized clustering regardless of the distribution of data among devices.",

keywords = "Blockchain, Collaborative Learning, Committee Consensus, Federated Learning, Privacy, Reputation",

author = "Mina Alishahi and Wouter Leeuw and Nicola Zannone",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 22nd IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2023 ; Conference date: 01-11-2023 Through 03-11-2023",

year = "2024",

month = may,

day = "29",

doi = "10.1109/TrustCom60117.2023.00184",

language = "English",

pages = "1348--1354",

editor = "Jia Hu and Geyong Min and Guojun Wang",

booktitle = "2023 IEEE 22nd International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom)",

publisher = "Institute of Electrical and Electronics Engineers",

address = "United States",

}

Alishahi, M, Leeuw, W & Zannone, N 2024, BC-FL k-means: A Blockchain-based Framework for Federated Clustering. in J Hu, G Min & G Wang (uitgave), 2023 IEEE 22nd International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom). Institute of Electrical and Electronics Engineers, blz. 1348-1354, 22nd IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2023, Exeter, Verenigd Koninkrijk, 1/11/23. https://doi.org/10.1109/TrustCom60117.2023.00184

BC-FL k-means: A Blockchain-based Framework for Federated Clustering. / Alishahi, Mina; Leeuw, Wouter; Zannone, Nicola.
2023 IEEE 22nd International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom). uitgave / Jia Hu; Geyong Min; Guojun Wang. Institute of Electrical and Electronics Engineers, 2024. blz. 1348-1354.

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/Congresprocedure › Conferentiebijdrage › Academic › peer review

TY - GEN

T1 - BC-FL k-means

T2 - 22nd IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2023

AU - Alishahi, Mina

AU - Leeuw, Wouter

AU - Zannone, Nicola

PY - 2024/5/29

Y1 - 2024/5/29

N2 - This work presents a novel framework to train clustering models collaboratively without compromising accuracy while accommodating privacy and security in a decentralized manner. Our decentralized collaborative learning model removes the single point of failure and excludes unreliable input by designing a committee-based consensus method in a blockchain-based federated learning, which is equipped with a reputation system. We present a prototype implementation of our approach and show that its performance is comparable with centralized clustering regardless of the distribution of data among devices.

AB - This work presents a novel framework to train clustering models collaboratively without compromising accuracy while accommodating privacy and security in a decentralized manner. Our decentralized collaborative learning model removes the single point of failure and excludes unreliable input by designing a committee-based consensus method in a blockchain-based federated learning, which is equipped with a reputation system. We present a prototype implementation of our approach and show that its performance is comparable with centralized clustering regardless of the distribution of data among devices.

KW - Blockchain

KW - Collaborative Learning

KW - Committee Consensus

KW - Federated Learning

KW - Privacy

KW - Reputation

UR - http://www.scopus.com/inward/record.url?scp=85195482425&partnerID=8YFLogxK

U2 - 10.1109/TrustCom60117.2023.00184

DO - 10.1109/TrustCom60117.2023.00184

M3 - Conference contribution

AN - SCOPUS:85195482425

SP - 1348

EP - 1354

BT - 2023 IEEE 22nd International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom)

A2 - Hu, Jia

A2 - Min, Geyong

A2 - Wang, Guojun

PB - Institute of Electrical and Electronics Engineers

Y2 - 1 November 2023 through 3 November 2023

ER -

BC-FL k-means: A Blockchain-based Framework for Federated Clustering

Samenvatting

Congres

Bibliografische nota

Toegang tot document

Andere bestanden en links

Vingerafdruk

Citeer dit