Abstract:Federated learning (FL) has become a prevalent distributed machine learning paradigm with improved privacy. After learning, the resulting federated model should be further personalized to each different client. While several methods have been proposed to achieve personalization, they are typically limited to a single local device, which may incur bias or overfitting since data in a single device is extremely limited. In this paper, we attempt to realize personalization beyond a single client. The motivation is that during the FL process, there may exist many clients with similar data distribution, and thus the personalization performance could be significantly boosted if these similar clients can cooperate with each other. Inspired by this, this paper introduces a new concept called federated adaptation, targeting at adapting the trained model in a federated manner to achieve better personalization results. However, the key challenge for federated adaptation is that we could not outsource any raw data from the client during adaptation, due to privacy concerns. In this paper, we propose PFA, a framework to accomplish Privacy-preserving Federated Adaptation. PFA leverages the sparsity property of neural networks to generate privacy-preserving representations and uses them to efficiently identify clients with similar data distributions. Based on the grouping results, PFA conducts an FL process in a group-wise way on the federated model to accomplish the adaptation. For evaluation, we manually construct several practical FL datasets based on public datasets in order to simulate both the class-imbalance and background-difference conditions. Extensive experiments on these datasets and popular model architectures demonstrate the effectiveness of PFA, outperforming other state-of-the-art methods by a large margin while ensuring user privacy. We will release our code at: https://github.com/lebyni/PFA.

Few-Shot Model Agnostic Federated Learning

Federated Few-Shot Learning with Adversarial Learning

Few-Shot Federated Learning: A Federated Learning Model for Small-Sample Scenarios

FewFedWeight: Few-shot Federated Learning Framework Across Multiple NLP Tasks

Personalized Federated Few-Shot Learning

Personalized Federated Learning on Long-Tailed Data Via Adversarial Feature Augmentation.

Personalized Federated Learning via Amortized Bayesian Meta-Learning

LF3PFL: A Practical Privacy-Preserving Federated Learning Algorithm Based on Local Federalization Scheme

Federated mutual learning

FLea: Addressing Data Scarcity and Label Skew in Federated Learning via Privacy-preserving Feature Augmentation

Client selection based weighted federated few-shot learning

MpFedcon : Model-Contrastive Personalized Federated Learning with the Class Center

FedMD: Heterogenous Federated Learning via Model Distillation

RAFLS: RDP-Based Adaptive Federated Learning with Shuffle Model

FedAA: Using Non-sensitive Modalities to Improve Federated Learning while Preserving Image Privacy

Sparse Federated Learning with Hierarchical Personalization Models.

Improving Generalization and Personalization in Model-Heterogeneous Federated Learning

FedAA: A Reinforcement Learning Perspective on Adaptive Aggregation for Fair and Robust Federated Learning

Fairness-aware Agnostic Federated Learning

Personalized federated learning based on feature fusion

PFA: Privacy-preserving Federated Adaptation for Effective Model Personalization