Towards Personalized Federated Learning

By Alysa Ziying Tan, Han Yu, Li-zhen Cui, Qiang Yang. IEEE Transactions on Neural Networks and Learning Systems, 2021

1,295 citations recorded by the source corpus.

In parallel with the rapid adoption of artificial intelligence (AI) empowered by advances in AI research, there has been growing awareness and concerns of data privacy. Recent significant developments in the data regulation landscape have prompted a seismic shift in interest toward privacy-preserving AI. This has contributed to the popularity of Federated Learning (FL), the leading paradigm for the training of machine learning models on data silos in a privacy-preserving manner. In this survey, we explore the domain of personalized FL (PFL) to address the fundamental challenges of FL on heterogeneous data, a universal characteristic inherent in all real-world datasets. We analyze the key motivations for PFL and present a unique taxonomy of PFL techniques categorized according to the key challenges and personalization strategies in PFL. We highlight their key ideas, challenges, opportunities, and envision promising future trajectories of research toward a new PFL architectural design, realistic PFL benchmarking, and trustworthy PFL approaches.

Subfields: Artificial Intelligence, Information Systems

View Paper: https://www.semanticscholar.org/paper/481dd25896ac531707870c9b8c179cce20013401

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Philip S. Yu
2. Guodong Long
3. Jing Jiang
4. Chengqi Zhang
5. Dacheng Tao
6. Qiang Yang
7. Tianyi Zhou
8. Bo Du

In Information Systems

1. Dusist Niyato
2. Zehui Xiong
3. Jiawen Kang
4. Bo Du
5. Mang Ye
6. Jing Jiang
7. Guodong Long
8. Tianyi Zhou

Citations in the corpus

• FedProto: Federated Prototype Learning across Heterogeneous Clients (2021, background): “Another widely studied strategy is personalized FL (Tan et al. 2021) where a personalized model is generated for each client by leveraging both global and local information.”
• Heterogeneous Federated Learning: State-of-the-art and Research Challenges (2023, background): “[198] They explore the field of personalized federated learning and conduct a taxonomic survey of existing methods.”
• Vertical Federated Learning: Concepts, Advances, and Challenges (2022, background): “While the applications and research on VFL have grown dramatically in recent years, there lacks a comprehensive survey on the advances, challenges, and potential research directions of VFL. Existing FL surveys focus either on HFL [6], [7], [14] or a limited perspective of VFL [3], [15].”
• Survey on Federated Learning Threats: concepts, taxonomy on attacks and defences, experimental study and challenges (2022, background): “For example, some of the most popular trends are Personalised FL [176] or fairness in FL [177].”
• A Full Dive Into Realizing the Edge-Enabled Metaverse: Visions, Enabling Technologies, and Challenges (2022, background): “In the food and beverage industry, AI-enabled HMDs can provide personalized AR/VR recommendations to users with regard to the users’ preferences [212], [213], [214].”
• FedBABU: Towards Enhanced Representation for Federated Image Classification (2021, background): “In addition, various technologies and algorithms for personalized FL have been presented and will be helpful to read; see Tan et al. (2021) and Kulkarni et al. (2020) for more details.”
• Decentralized Federated Learning: A Survey and Perspective (2023, background): “…can be more personalized to the state’s user scenarios and user habits, such as weather, number of users, time zone, ethnicity, age distribution, etc. 2) Personalized FL can be classified into two categories, i.e., global model personalization and personalized model architecture [120], [121], [122].”
• Federated learning for medical image analysis: A survey (2023, methodology): “Another promising solution is personalized FL techniques (T Dinh et al., 2020; Tan et al., 2022) which utilize local data to further optimize a trained global model.”
• Agentic AI: Autonomous Intelligence for Complex Goals—A Comprehensive Survey (2025, neutral): “There is ongoing advancement in deploying distributed AI frameworks, such as federated learning [81] and edge AI [82], that enhance the maneuverability of Agentic AI systems in a wider operational area with dispersed installations [83].”