Straggler-Resilient Personalized Federated Learning
Abstract: Federated Learning is an emerging learning paradigm that allows training models from samples distributed across a large network of clients while respecting privacy and communication restrictions. Despite its success, federated learning faces several challenges related to its decentralized nature. In this work, we develop a novel algorithmic procedure with theoretical speedup guarantees that simultaneously handles two of these hurdles, namely (i) data heterogeneity, i.e., data distributions can vary substantially across clients, and (ii) system heterogeneity, i.e., the computational power of the clients could differ significantly. By leveraging previous works in the realm of representation learning (Collins et al., 2021; Liang et al., 2020), our method constructs a global common representation utilizing the data from all clients. Additionally, it learns a user-specific set of parameters resulting in a personalized solution for each individual client. Furthermore, it mitigates the effects of stragglers by adaptively selecting clients based on their computational characteristics, thus achieving for the first time near optimal sample complexity and provable logarithmic speedup. Experimental results support our theoretical findings showing the superiority of our method over alternative personalized federated schemes in system and data heterogeneous environments.
License: Creative Commons Attribution 4.0 International (CC BY 4.0)
Submission Length: Regular submission (no more than 12 pages of main content)
Code: https://github.com/shenzebang/SRPFL https://github.com/shenzebang/PyTorch-Sentiment-Analysis
Supplementary Material: zip
Assigned Action Editor: ~Virginia_Smith1
Submission Number: 1184