Awesome-Personalized-Federated-Learning

Paper List for Personalized Federated Learning (PFL)

Personalized Federated Learning (PFL), allowing different parameters among clients, is designed for data heterogeneity (statistics heterogeneity) problem in federated learning.

We categorized PFL algorithms into four types and collected corresponding papers

• Parameters-based: These algorithms emphasize personalization through their model parameters, enabling the customization of parameters within specific parts of the model.
• Full-personalization-based: These algorithms tailor a personalized model for each client. The global information is injected via techniques like mix-up or global regularization.
• Similarity-based: These algorithms focus on capturing the similarities among clients and aim to facilitate synergy between similar clients.
• Learn-to-personalize-based: These algorithms are designed to identify methods for easily personalizing models locally.

A detailed mindmap of paper collected is illustrated as follow

We also collect PFL papers like

• Per-instance PFL
• PFL with Transformers
• Out-of-distribution (OOD) in PFL

More items will be added to the repository. Please feel free to suggest other key resources by opening an issue report, submitting a pull request, or dropping me an email @ ([email protected]). 🥳

0. Survey & Benchmark

Survey & Benchmark

Title	Abbreviation	Venue	Year	Materials
PFLlib: Personalized Federated Learning Algorithm Library	PFLlib	arXiv	2023	[PDF] [CODE]
Towards Personalized Federated Learning	N/A	TNNLS	2022	[PDF]
pFL-Bench: A Comprehensive Benchmark for Personalized Federated Learning	pFL-Bench	NIPS	2022	[PDF] [CODE]
Motley: Benchmarking heterogeneity and personalization in federated learning	Motley	arXiv	2022	[PDF] [CODE]
Federated Machine Learning: Concept and Applications	N/A	TIST	2019	[PAGE]

1. Parameters

Split Learning

Title	Abbreviation	Venue	Year	Materials
Improving Global Generalization and Local Personalization for Federated Learning	pFedCSPC	TNNLS	2024	[PAGE]
Test-Time Robust Personalization for Federated Learning	FedTHE	ICLR	2023	[PDF] [CODE]
Personalized Federated Learning with Feature Alignment and Classifier Collaboration	FedPAC	ICLR	2023	[PDF] [CODE]
FedCP: Separating Feature Information for Personalized Federated Learning via Conditional Policy	FedCP	KDD	2023	[PAGE] [CODE]
On bridging generic and personalized federated learning for image classification	FedRoD	ICLR	2022	[PDF] [CODE]
Fedbabu: Towards enhanced representation for federated image classification	Fedbabu	ICLR	2022	[PDF] [CODE]
Exploiting shared representations for personalized federated learning	FedRep	ICML	2021	[PDF] [CODE]
Adapt to Adaptation: Learning Personalization for Cross-Silo Federated Learning	APPLE	IJCAI	2021	[PDF] [CODE]
Think locally, act globally: Federated learning with local and global representations	LG-FedAvg	arXiv	2020	[PDF] [CODE]
Federated Learning with Personalization Layers	FedPer	arXiv	2019	[PDF] [CODE]

Distillation

Title	Abbreviation	Venue	Year	Materials
Spectral Co-Distillation for Personalized Federated Learning	PFL+	NIPS	2023	[PDF]
Cd2-pfed: Cyclic distillation-guided channel decoupling for model personalization in federated learning	CD2-pFed	CVPR	2022	[PDF]
Quped: Quantized personalization via distillation with applications to federated learning	QuPeD	NIPS	2021	[PDF]
Data-Free Knowledge Distillation for Heterogeneous Federated Learning	FedGEN	ICML	2021	[PDF]
Federated model distillation with noise-free differential privacy	FedMD-NFDP	IJCAI	2021	[PDF]
Parameterized Knowledge Transfer for Personalized Federated Learning	KT-pFL	NIPS	2021	[PDF]
Ensemble distillation for robust model fusion in federated learning	FedDF	NIPS	2020	[PDF]
Fedmd: Heterogenous federated learning via model distillation	FedMD	arXiv	2019	[PDF]
Communication-efficient on-device machine learning: Federated distillation and augmentation under non-iid private data	FD	arXiv	2018	[PDF]

Sparsity

Title	Abbreviation	Venue	Year	Materials
FedP3: Federated Personalized and Privacy-friendly Network Pruning under Model Heterogeneity	FedP3	ICLR	2024	[PDF]
DM-PFL: Hitchhiking Generic Federated Learning for Efficient Shift-Robust Personalization	DM-PFL	KDD	2023	[PAGE]
Efficient Personalized Federated Learning via Sparse Model-Adaptation	pFedGate	ICML	2023	[PDF] [CODE]
PerFedMask: Personalized federated learning with optimized masking vectors	PerFedMask	ICLR	2023	[PDF] [CODE]
Personalized Subgraph Federated Learning	FED-PUB	ICML	2023	[PDF] [CODE]
Dynamic Personalized Federated Learning with Adaptive Differential Privacy	FedDPA	NIPS	2023	[PDF] [CODE]
DisPFL: Towards Communication-Efficient Personalized Federated Learning via Decentralized Sparse Training	DisPFL	ICML	2022	[PAGE] [CODE]
Tailorfl: Dual-personalized federated learning under system and data heterogeneity	TailorFL	SenSys	2022	[PDF]
FedMask: Joint Computation and Communication-Efficient Personalized Federated Learning via Heterogeneous Masking	FedMask	SenSys	2021	[PDF]
Hermes: an efficient federated learning framework for heterogeneous mobile clients	Hermes	MobiCom	2021	[PAGE]

2. Full-Personalization

Mix-up

Title	Abbreviation	Venue	Year	Materials
Heterogeneous Personalized Federated Learning by Local-Global Updates Mixing via Convergence Rate	LG-Mix	ICLR	2024	[PDF] [CODE]
FedALA: Adaptive Local Aggregation for Personalized Federated Learning	FedALA	AAAI	2023	[PAGE] [CODE]
Adaptive personalized federated learning	APFL	arXiv	2020	[PDF]

Global Regularization

Title	Abbreviation	Venue	Year	Materials
Ditto: Fair and Robust Federated Learning Through Personalization	Ditto	ICML	2021	[PDF] [CODE]
Personalized federated learning with moreau envelopes	pFedMe	NIPS	2020	[PDF] [CODE]

3. Similarity

Clustering

Title	Abbreviation	Venue	Year	Materials
CASA: Clsutered Federated Learning for Asynchronous Clients	CASA	KDD	2024
EchoPFL: Asynchronous Personalized Federated Learning on Mobile Devices with On-Demand Staleness Control	EchoPFL	UbiComp	2024	[PAGE] [PDF]
Exploiting Label Skews in Federated Learning with Model Concatenation	FedConcat	AAAI	2024	[PAGE]
Adaptive Group Personalization for Federated Mutual Transfer Learning	AdaGrp	ICML	2024
Clustered Federated Learning via Gradient Partitioning	N/A	ICML	2024
FedRC: Tackling Diverse Distribution Shifts Challenge in Federated Learning by Robust Clustering	FedRC	ICML	2024
Clustered Federated Learning in Heterogeneous Environment	ICFL	TNNLS	2023	[PAGE]
Auxo: Efficient Federated Learning via Scalable Client Clustering	Auxo	SoCC	2023	[PAGE] [PDF]
Multi-center federated learning: clients clustering for better personalization	FeSEM	WWW	2023	[PAGE]
ClusterFL: A Similarity-Aware Federated Learning Systemfor Human Activity Recognition	ClusterFL	MobiSys	2021	[PAGE] [CODE]
DistFL: Distribution-aware Federated Learning for Mobile Scenarios	DistFL	UbiComp	2021	[PAGE] [PDF]
Clustered Federated Learning: Model-Agnostic Distributed Multitask Optimization Under Privacy Constraints	CFL	TNNLS	2020	[PDF] [CODE]
An Efficient Framework for Clustered Federated Learning	IFCA	NIPS	2020	[PDF] [CODE]
Federated learning with hierarchical clustering of local updates to improve training on non-IID data	FL+HC	IJCNN	2020	[PAGE] [PDF]

Collaboration

Title	Abbreviation	Venue	Year	Materials
Rethinking Personalized Client Collaboration in Federated Learning	pFedSV	TMC	2024	[PAGE]
Personalized Federated Learning with Inferred Collaboration Graphs	pFedGraph	ICML	2023	[PDF] [CODE]
Personalized Federated Learning with Feature Alignment and Classifier Collaboration	FedPAC	ICLR	2023	[PDF] [CODE]
Personalized Cross-Silo Federated Learning on Non-IID Data	FedAMP	AAAI	2021	[PDF]
Variational federated multi-task learning	VIRTUAL	arXiv	2019	[PDF]
Federated Multi-Task Learning	MOCHA	NIPS	2017	[PDF] [CODE]

4. Learn to Personalize

Meta Learning

Title	Abbreviation	Venue	Year	Materials
Personalized Federated Learning with Parameter Propagation	FEDORA	KDD	2023	[PAGE]
FedL2P: Federated Learning to Personalize	FedL2P	NIPS	2023	[PDF] [CODE]
Personalized Federated Learning with Theoretical Guarantees: A Model-Agnostic Meta-Learning Approach	Per-FedAvg	NIPS	2020	[PDF]
Improving federated learning personalization via model agnostic meta learning	MAML	arXiv	2019	[PDF]

Hyper-network

Title	Abbreviation	Venue	Year	Materials
FedMBridge: Bridgeable Multimodal Federated Learning	FedMBridge	ICML	2024	[PDF]
PeFLL: Personalized Federated Learning by Learning to Learn	PeFLL	ICLR	2023	[PDF]
Personalized federated learning using hypernetworks	pFedHN	ICML	2021	[PDF] [CODE]

Test-time Adaptation

Title	Abbreviation	Venue	Year	Materials
Adaptive Test-Time Personalization for Federated Learning	ATP	NIPS	2023	[PDF] [CODE]

5. Other

Per-instance PFL

Title	Abbreviation	Venue	Year	Materials
Flow: Per-instance Personalized Federated Learning	Flow	NIPS	2023	[PDF] [CODE]
FedCP: Separating Feature Information for Personalized Federated Learning via Conditional Policy	FedCP	KDD	2023	[PAGE] [CODE]
Personalized Federated Learning through Local Memorization	KNN-Per	ICML	2022	[PDF] [CODE]

PFL in Transformers

Title	Abbreviation	Venue	Year	Materials
FedPerfix: Towards Partial Model Personalization of Vision Transformers in Federated Learning	FedPerfix	ICCV	2023	[PDF] [CODE]
FedTP: Federated Learning by Transformer Personalization	FedTP	TNNLS	2023	[PAGE] [PDF]
Efficient Model Personalization in Federated Learning via Client-Specific Prompt Generation	pFedPG	ICCV	2023	[PDF]

Out-of-distribution/Generalization in PFL

Title	Abbreviation	Venue	Year	Materials
Learning Personalized Causally Invariant Representations for Heterogeneous Federated Clients	FedSDR	ICLR	2024	[PDF] [CODE]
Improving Global Generalization and Local Personalization for Federated Learning	pFedCSPC	TNNLS	2024	[PAGE]
Test-Time Robust Personalization for Federated Learning	FedTHE	ICLR	2023	[PDF] [CODE]
DM-PFL: Hitchhiking Generic Federated Learning for Efficient Shift-Robust Personalization	DM-PFL	KDD	2023	[PAGE]
On bridging generic and personalized federated learning for image classification	FedRoD	ICLR	2022	[PDF] [CODE]

Last Update: Jul 23, 2024 22:37:28