Paper Overview
Field: Machine Learning Authors: Abdulmoneam Ali, Ahmed Arafa Published: 2026-04-21 arXiv: 2604.19729
Abstract (translated)
Personalized Federated Learning (PFL) aims to learn multiple task-specific models rather than a single global model across heterogeneous data distributions. Existing PFL approaches typically rely on iterative optimization—such as model update trajectories—to cluster users that need to accomplish the same tasks together. However, these learning-dynamics-based methods are inherently vulnerable to low-quality data and noisy labels, as corrupted updates distort clustering decisions and degrade personalization performance.
To tackle this, the authors propose FB-NLL, a feature-centric framework that decouples user clustering from iterative training dynamics. By exploiting the intrinsic heterogeneity of local feature spaces, FB-NLL characterizes each user through the spectral structure of the covariances of their feature representations, and uses subspace similarity to identify task-consistent user groupings. Key properties:
- Geometry-aware clustering that is label-agnostic and performed in a one-shot manner before training, significantly reducing communication overhead and computation cost.
- Feature-consistency-based noise handling: a detection and correction strategy leverages directional alignment in the learned feature space and assigns labels based on class-specific feature subspaces, mitigating corrupted supervision without estimating stochastic noise transition matrices.
- Model-agnostic design that integrates seamlessly with existing noise-robust training techniques.
- arXiv: https://arxiv.org/abs/2604.19729
Extensive experiments across diverse datasets and noise mechanisms show that the framework consistently outperforms state-of-the-art baselines in both average accuracy and performance stability.