Title:
METHOD AND SYSTEM FOR FEDERATED LEARNING
Document Type and Number:
WIPO Patent Application WO/2024/072074
Kind Code:
A1
Abstract:
Broadly speaking, embodiments of the present techniques provide a method for training a machine learning, ML, model using a server and a plurality of client devices using federated learning, FL. Advantageously, the present FL method does not require access to user data on the client devices, and does not require repeatedly sharing large amounts of data (e.g. full models) between the server and client devices. Furthermore, the present FL method enables a large variety of client devices to participate in the FL, which reduces bias in the trained ML model.
More Like This:
Inventors:
LI DA (GB)
HU XU (GB)
HOSPEDALES TIMOTHY (GB)
HU XU (GB)
HOSPEDALES TIMOTHY (GB)
Application Number:
PCT/KR2023/014974
Publication Date:
April 04, 2024
Filing Date:
September 27, 2023
Export Citation:
Assignee:
SAMSUNG ELECTRONICS CO LTD (KR)
International Classes:
G06N3/098; G06N3/0455; G06N3/096
Foreign References:
| US20220083904A1 | 2022-03-17 |
Other References:
DONGQI CAI: "FedAdapter: Efficient Federated Learning for Modern NLP", ARXIV (CORNELL UNIVERSITY), CORNELL UNIVERSITY LIBRARY, ARXIV.ORG, ITHACA, 8 May 2023 (2023-05-08), Ithaca, XP093152160, Retrieved from the Internet DOI: 10.48550/arxiv.2205.10162
JAE RO: "Scaling Language Model Size in Cross-Device Federated Learning", PROCEEDINGS OF THE FIRST WORKSHOP ON FEDERATED LEARNING FOR NATURAL LANGUAGE PROCESSING (FL4NLP 2022), ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, STROUDSBURG, PA, USA, 24 June 2022 (2022-06-24), Stroudsburg, PA, USA, pages 6 - 20, XP093152162, DOI: 10.18653/v1/2022.fl4nlp-1.2
TIEN-JU YANG: "Partial Variable Training for Efficient on-Device Federated Learning", ICASSP 2022, 2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), IEEE, 23 May 2022 (2022-05-23) - 27 May 2022 (2022-05-27), pages 4348 - 4352, XP093152165, ISBN: 978-1-6654-0540-9, DOI: 10.1109/ICASSP43922.2022.9746836
HAKIM SIDAHMED; ZHENG XU; ANKUSH GARG; YUAN CAO; MINGQING CHEN: "Efficient and Private Federated Learning with Partially Trainable Networks", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, 6 October 2021 (2021-10-06), 201 Olin Library Cornell University Ithaca, NY 14853, XP091072325
JAE RO: "Scaling Language Model Size in Cross-Device Federated Learning", PROCEEDINGS OF THE FIRST WORKSHOP ON FEDERATED LEARNING FOR NATURAL LANGUAGE PROCESSING (FL4NLP 2022), ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, STROUDSBURG, PA, USA, 24 June 2022 (2022-06-24), Stroudsburg, PA, USA, pages 6 - 20, XP093152162, DOI: 10.18653/v1/2022.fl4nlp-1.2
TIEN-JU YANG: "Partial Variable Training for Efficient on-Device Federated Learning", ICASSP 2022, 2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP), IEEE, 23 May 2022 (2022-05-23) - 27 May 2022 (2022-05-27), pages 4348 - 4352, XP093152165, ISBN: 978-1-6654-0540-9, DOI: 10.1109/ICASSP43922.2022.9746836
HAKIM SIDAHMED; ZHENG XU; ANKUSH GARG; YUAN CAO; MINGQING CHEN: "Efficient and Private Federated Learning with Partially Trainable Networks", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, 6 October 2021 (2021-10-06), 201 Olin Library Cornell University Ithaca, NY 14853, XP091072325
Attorney, Agent or Firm:
KIM, Tae-hun et al. (KR)
Download PDF: