Poster Abstract: OrderlessFL: A CRDT-Enabled Permissioned Blockchain for Federated Learning

Pezhman Nasirifard; Ruben Mayer; Hans Arno Jacobsen

doi:10.1145/3565386.3565487

Poster Abstract: OrderlessFL: A CRDT-Enabled Permissioned Blockchain for Federated Learning

Pezhman Nasirifard, Ruben Mayer, Hans Arno Jacobsen

Informatics 13 - Chair of Decentralized Information Systems and Data Management

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

1 Scopus citations

Abstract

Industries produce a large amount of data that can improve Machine Learning models. However, due to privacy issues, the data cannot be shared. Several Federated Learning (FL) systems have been introduced as private alternatives without considering Byzantine actors. Also, these systems are affected by the gradient staleness problem. Several blockchain-based FL systems are introduced to address Byzantine actors, which rely on Proof-of-Work-based (PoW) protocols and suffer from their limitations. We introduce OrderlessFL, a safe permissioned blockchain-based FL system using flCRDT, a CRDT for concurrent ML training and mitigating gradient staleness.

Original language	English
Title of host publication	Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022
Publisher	Association for Computing Machinery, Inc
Pages	7-8
Number of pages	2
ISBN (Electronic)	9781450399319
DOIs	https://doi.org/10.1145/3565386.3565487
State	Published - 7 Nov 2022
Event	23rd International Middleware Conference, Middleware 2022 - Part of Middleware 2022 - Quebec, Canada Duration: 7 Nov 2022 → 11 Nov 2022

Publication series

Name	Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022

Conference

Conference	23rd International Middleware Conference, Middleware 2022 - Part of Middleware 2022
Country/Territory	Canada
City	Quebec
Period	7/11/22 → 11/11/22

Keywords

blockchain
conflict-free replicated data type
federated learning

Access to Document

10.1145/3565386.3565487

Cite this

Nasirifard, P., Mayer, R., & Jacobsen, H. A. (2022). Poster Abstract: OrderlessFL: A CRDT-Enabled Permissioned Blockchain for Federated Learning. In Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022 (pp. 7-8). (Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022). Association for Computing Machinery, Inc. https://doi.org/10.1145/3565386.3565487

Nasirifard, Pezhman ; Mayer, Ruben ; Jacobsen, Hans Arno. / Poster Abstract : OrderlessFL: A CRDT-Enabled Permissioned Blockchain for Federated Learning. Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022. Association for Computing Machinery, Inc, 2022. pp. 7-8 (Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022).

@inproceedings{252f66e6ad1e42f8bb10f2ec9a80ad45,

title = "Poster Abstract: OrderlessFL: A CRDT-Enabled Permissioned Blockchain for Federated Learning",

abstract = "Industries produce a large amount of data that can improve Machine Learning models. However, due to privacy issues, the data cannot be shared. Several Federated Learning (FL) systems have been introduced as private alternatives without considering Byzantine actors. Also, these systems are affected by the gradient staleness problem. Several blockchain-based FL systems are introduced to address Byzantine actors, which rely on Proof-of-Work-based (PoW) protocols and suffer from their limitations. We introduce OrderlessFL, a safe permissioned blockchain-based FL system using flCRDT, a CRDT for concurrent ML training and mitigating gradient staleness.",

keywords = "blockchain, conflict-free replicated data type, federated learning",

author = "Pezhman Nasirifard and Ruben Mayer and Jacobsen, {Hans Arno}",

note = "Publisher Copyright: {\textcopyright} 2022 Owner/Author.; 23rd International Middleware Conference, Middleware 2022 - Part of Middleware 2022 ; Conference date: 07-11-2022 Through 11-11-2022",

year = "2022",

month = nov,

day = "7",

doi = "10.1145/3565386.3565487",

language = "English",

series = "Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022",

publisher = "Association for Computing Machinery, Inc",

pages = "7--8",

booktitle = "Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022",

}

Nasirifard, P, Mayer, R & Jacobsen, HA 2022, Poster Abstract: OrderlessFL: A CRDT-Enabled Permissioned Blockchain for Federated Learning. in Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022. Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022, Association for Computing Machinery, Inc, pp. 7-8, 23rd International Middleware Conference, Middleware 2022 - Part of Middleware 2022, Quebec, Canada, 7/11/22. https://doi.org/10.1145/3565386.3565487

Poster Abstract: OrderlessFL: A CRDT-Enabled Permissioned Blockchain for Federated Learning. / Nasirifard, Pezhman; Mayer, Ruben ; Jacobsen, Hans Arno.
Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022. Association for Computing Machinery, Inc, 2022. p. 7-8 (Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022).

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

TY - GEN

T1 - Poster Abstract

T2 - 23rd International Middleware Conference, Middleware 2022 - Part of Middleware 2022

AU - Nasirifard, Pezhman

AU - Mayer, Ruben

AU - Jacobsen, Hans Arno

PY - 2022/11/7

Y1 - 2022/11/7

N2 - Industries produce a large amount of data that can improve Machine Learning models. However, due to privacy issues, the data cannot be shared. Several Federated Learning (FL) systems have been introduced as private alternatives without considering Byzantine actors. Also, these systems are affected by the gradient staleness problem. Several blockchain-based FL systems are introduced to address Byzantine actors, which rely on Proof-of-Work-based (PoW) protocols and suffer from their limitations. We introduce OrderlessFL, a safe permissioned blockchain-based FL system using flCRDT, a CRDT for concurrent ML training and mitigating gradient staleness.

AB - Industries produce a large amount of data that can improve Machine Learning models. However, due to privacy issues, the data cannot be shared. Several Federated Learning (FL) systems have been introduced as private alternatives without considering Byzantine actors. Also, these systems are affected by the gradient staleness problem. Several blockchain-based FL systems are introduced to address Byzantine actors, which rely on Proof-of-Work-based (PoW) protocols and suffer from their limitations. We introduce OrderlessFL, a safe permissioned blockchain-based FL system using flCRDT, a CRDT for concurrent ML training and mitigating gradient staleness.

KW - blockchain

KW - conflict-free replicated data type

KW - federated learning

UR - http://www.scopus.com/inward/record.url?scp=85145584588&partnerID=8YFLogxK

U2 - 10.1145/3565386.3565487

DO - 10.1145/3565386.3565487

M3 - Conference contribution

AN - SCOPUS:85145584588

T3 - Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022

SP - 7

EP - 8

BT - Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022

PB - Association for Computing Machinery, Inc

Y2 - 7 November 2022 through 11 November 2022

ER -

Nasirifard P, Mayer R , Jacobsen HA. Poster Abstract: OrderlessFL: A CRDT-Enabled Permissioned Blockchain for Federated Learning. In Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022. Association for Computing Machinery, Inc. 2022. p. 7-8. (Middleware 2022 - Proceedings of the 23rd International Middleware Conference Demos and Posters, Part of Middleware 2022). doi: 10.1145/3565386.3565487

Poster Abstract: OrderlessFL: A CRDT-Enabled Permissioned Blockchain for Federated Learning

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this