Abstract
By allowing users to erase their data's impact on federated learning models, federated unlearning protects users' right to be forgotten and data privacy. Despite a burgeoning body of research on federated unlearning's technical feasibility, there is a paucity of literature investigating the considerations behind users' requests for data revocation. This paper proposes a non-cooperative game framework to study users' data revocation strategies in federated unlearning. We prove the existence of a Nash equilibrium. However, users' best response strategies are coupled via model performance and unlearning costs, which makes the equilibrium computation challenging. We obtain the Nash equilibrium by establishing its equivalence with a much simpler auxiliary optimization problem. We also summarize users' multi-dimensional attributes into a single-dimensional metric and derive the closed-form characterization of an equilibrium, when users' unlearning costs are negligible. Moreover, we compare the cases of allowing and forbidding partial data revocation in federated unlearning. Interestingly, the results reveal that allowing partial revocation does not necessarily increase users' data contributions or payoffs due to the game structure. Additionally, we demonstrate that positive externalities may exist between users' data revocation decisions when users incur unlearning costs, while this is not the case when their unlearning costs are negligible.
Original language | English (US) |
---|---|
Title of host publication | IEEE INFOCOM 2024 - IEEE Conference on Computer Communications |
Publisher | Institute of Electrical and Electronics Engineers Inc. |
Pages | 1151-1160 |
Number of pages | 10 |
ISBN (Electronic) | 9798350383508 |
DOIs | |
State | Published - 2024 |
Event | 2024 IEEE Conference on Computer Communications, INFOCOM 2024 - Vancouver, Canada Duration: May 20 2024 → May 23 2024 |
Publication series
Name | Proceedings - IEEE INFOCOM |
---|---|
ISSN (Print) | 0743-166X |
Conference
Conference | 2024 IEEE Conference on Computer Communications, INFOCOM 2024 |
---|---|
Country/Territory | Canada |
City | Vancouver |
Period | 5/20/24 → 5/23/24 |
Funding
This work is supported by NSF ECCS-2030251 and NSF ECCS-2216970.
Keywords
- allowing or forbidding partial revocation
- data revocation strategies
- Federated unlearning
- game theory
ASJC Scopus subject areas
- General Computer Science
- Electrical and Electronic Engineering