f-FUM: Federated Unlearning via min--max and f-divergence
cs.LG
Submitted: 2026-02-05
Updated: 2026-09-26
License: http://creativecommons.org/licenses/by/4.0/
The gist: Federated learning (FL) enables collaborative training while keeping raw data on clients.
Terminology
Abstract
Federated learning (FL) enables collaborative training while keeping raw data on clients. Deletion requests and the discovery of poisoned data create a need to remove selected contributions from an already trained model. This is challenging in FL because client data are decentralized and their influence is entangled through repeated aggregation. We present f-FUM, an active federated unlearning method that builds on teacher-student forget/retain optimization. Starting from a pretrained global model, the method keeps that model fixed as a teacher. Clients maximize an f-divergence between student and teacher predictive distributions on forget examples, then minimize KL-based teacher-student disagreement and supervised loss on retained examples. The updates are aggregated in separate, sample-weighted synchronization phases without moving raw data to the server or changing the model architecture. We evaluate forget-side divergence choices in client-level and data-level deletion settings involving backdoors, label confusion, and clean-data deletion. Under equal synchronization-phase budgets, divergence choice changes the forgetting-utility trade-off, yielding improvements in some settings and mixed results in others. In the evaluated configurations, f-FUM uses up to 8 times fewer synchronization phases than full federated retraining.
Sources
- Robust Semi-supervised Learning via $f$-Divergence and $\alpha$-R'enyi Divergence
- Flower: A Friendly Federated Learning Research Framework
- QuickDrop: Efficient Federated Unlearning by Integrated Dataset Distillation
- Understanding Fine-tuning in Approximate Unlearning: A Theoretical Perspective
- Attribute-to-Delete: Machine Unlearning via Datamodel Matching
- Generative Adversarial Networks
- Unlearning during Learning: An Efficient Federated Machine Unlearning Method
- BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain
- Certified Data Removal from Machine Learning Models
- Inexact Unlearning Needs More Careful Evaluations to Avoid a False Sense of Privacy
- An Information Theoretic Evaluation Metric For Strong Unlearning
- Towards Efficient and Certified Recovery from Poisoning Attacks in Federated Learning
- Forgettable Federated Linear Learning with Certified Data Unlearning
- NoT: Federated Unlearning via Weight Negation
- Towards Unbounded Machine Unlearning
- Subspace based Federated Unlearning
- Machine Unlearning: Taxonomy, Metrics, Applications, Challenges, and Prospects
- Class-wise Federated Unlearning: Harnessing Active Forgetting with Teacher-Student Memory Generation
- $f$-Divergence Based Classification: Beyond the Use of Cross-Entropy
- Learning to Forget using Hypernetworks
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks