SIFU: Sequential Informed Federated Unlearning for Efficient and Provable Client Unlearning in Federated Optimization

Yann Fraboni, Martin Van Waerebeke, Kevin Scaman, Richard Vidal, Laetitia Kameni, Marco Lorenzi
Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, PMLR 238:3457-3465, 2024.

Abstract

Machine Unlearning (MU) is an increasingly important topic in machine learning safety, aiming at removing the contribution of a given data point from a training procedure. Federated Unlearning (FU) consists in extending MU to unlearn a given client’s contribution from a federated training routine. While several FU methods have been proposed, we currently lack a general approach providing formal unlearning guarantees to the FedAvg routine, while ensuring scalability and generalization beyond the convex assumption on the clients’ loss functions. We aim at filling this gap by proposing SIFU (Sequential Informed Federated Unlearning), a new FU method applying to both convex and non-convex optimization regimes. SIFU naturally applies to FedAvg without additional computational cost for the clients and provides formal guarantees on the quality of the unlearning task. We provide a theoretical analysis of the unlearning properties of SIFU, and practically demonstrate its effectiveness as compared to a panel of unlearning methods from the state-of-the-art.

Cite this Paper

BibTeX
@InProceedings{pmlr-v238-fraboni24a, title = {{SIFU}: Sequential Informed Federated Unlearning for Efficient and Provable Client Unlearning in Federated Optimization}, author = {Fraboni, Yann and Van Waerebeke, Martin and Scaman, Kevin and Vidal, Richard and Kameni, Laetitia and Lorenzi, Marco}, booktitle = {Proceedings of The 27th International Conference on Artificial Intelligence and Statistics}, pages = {3457--3465}, year = {2024}, editor = {Dasgupta, Sanjoy and Mandt, Stephan and Li, Yingzhen}, volume = {238}, series = {Proceedings of Machine Learning Research}, month = {02--04 May}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v238/fraboni24a/fraboni24a.pdf}, url = {https://proceedings.mlr.press/v238/fraboni24a.html}, abstract = {Machine Unlearning (MU) is an increasingly important topic in machine learning safety, aiming at removing the contribution of a given data point from a training procedure. Federated Unlearning (FU) consists in extending MU to unlearn a given client’s contribution from a federated training routine. While several FU methods have been proposed, we currently lack a general approach providing formal unlearning guarantees to the FedAvg routine, while ensuring scalability and generalization beyond the convex assumption on the clients’ loss functions. We aim at filling this gap by proposing SIFU (Sequential Informed Federated Unlearning), a new FU method applying to both convex and non-convex optimization regimes. SIFU naturally applies to FedAvg without additional computational cost for the clients and provides formal guarantees on the quality of the unlearning task. We provide a theoretical analysis of the unlearning properties of SIFU, and practically demonstrate its effectiveness as compared to a panel of unlearning methods from the state-of-the-art.} }
Endnote
%0 Conference Paper %T SIFU: Sequential Informed Federated Unlearning for Efficient and Provable Client Unlearning in Federated Optimization %A Yann Fraboni %A Martin Van Waerebeke %A Kevin Scaman %A Richard Vidal %A Laetitia Kameni %A Marco Lorenzi %B Proceedings of The 27th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2024 %E Sanjoy Dasgupta %E Stephan Mandt %E Yingzhen Li %F pmlr-v238-fraboni24a %I PMLR %P 3457--3465 %U https://proceedings.mlr.press/v238/fraboni24a.html %V 238 %X Machine Unlearning (MU) is an increasingly important topic in machine learning safety, aiming at removing the contribution of a given data point from a training procedure. Federated Unlearning (FU) consists in extending MU to unlearn a given client’s contribution from a federated training routine. While several FU methods have been proposed, we currently lack a general approach providing formal unlearning guarantees to the FedAvg routine, while ensuring scalability and generalization beyond the convex assumption on the clients’ loss functions. We aim at filling this gap by proposing SIFU (Sequential Informed Federated Unlearning), a new FU method applying to both convex and non-convex optimization regimes. SIFU naturally applies to FedAvg without additional computational cost for the clients and provides formal guarantees on the quality of the unlearning task. We provide a theoretical analysis of the unlearning properties of SIFU, and practically demonstrate its effectiveness as compared to a panel of unlearning methods from the state-of-the-art.
APA
Fraboni, Y., Van Waerebeke, M., Scaman, K., Vidal, R., Kameni, L. & Lorenzi, M.. (2024). SIFU: Sequential Informed Federated Unlearning for Efficient and Provable Client Unlearning in Federated Optimization. Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 238:3457-3465 Available from https://proceedings.mlr.press/v238/fraboni24a.html.

Related Material