$f$-FUM: Federated Unlearning via min--max and $f$-divergence

TL;DR

This paper introduces a federated unlearning framework using a min-max optimization approach based on $f$-divergence, enabling efficient removal of data contributions while preserving model utility in decentralized settings.

Contribution

It proposes a novel federated unlearning method formulated as a min-max optimization problem that can be integrated into existing federated learning systems, unlike prior approaches.

Findings

Achieves significant speedups over naive retraining.

Maintains minimal utility degradation.

Effective in removing specific data contributions.

Abstract

Federated Learning (FL) has emerged as a powerful paradigm for collaborative machine learning across decentralized data sources, preserving privacy by keeping data local. However, increasing legal and ethical demands, such as the "right to be forgotten", and the need to mitigate data poisoning attacks have underscored the urgent necessity for principled data unlearning in FL. Unlike centralized settings, the distributed nature of FL complicates the removal of individual data contributions. In this paper, we propose a novel federated unlearning framework formulated as a min-max optimization problem, where the objective is to maximize an $f$-divergence between the model trained with all data and the model retrained without specific data points, while minimizing the degradation on retained data. Our framework could act like a plugin and be added to almost any federated setup, unlike SOTA methods like (\cite{10269017} which requires model degradation in server, or \cite{khalil2025notfederatedunlearningweight} which requires to involve model architecture and model weights). This formulation allows for efficient approximation of data removal effects in a federated setting. We provide empirical evaluations to show that our method achieves significant speedups over naive retraining, with minimal impact on utility.