Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2407.09360 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914867812761600 |
|---|---|
| author | Gu, Endong Chen, Yongxin Wen, Hao Cai, Xingju Han, Deren |
| author_facet | Gu, Endong Chen, Yongxin Wen, Hao Cai, Xingju Han, Deren |
| contents | This paper proposes LCFL, a novel clustering metric for evaluating clients' data distributions in federated learning. LCFL aligns with federated learning requirements, accurately assessing client-to-client variations in data distribution. It offers advantages over existing clustered federated learning methods, addressing privacy concerns, improving applicability to non-convex models, and providing more accurate classification results. LCFL does not require prior knowledge of clients' data distributions. We provide a rigorous mathematical analysis, demonstrating the correctness and feasibility of our framework. Numerical experiments with neural network instances highlight the superior performance of LCFL over baselines on several clustered federated learning benchmarks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_09360 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Novel clustered federated learning based on local loss Gu, Endong Chen, Yongxin Wen, Hao Cai, Xingju Han, Deren Machine Learning Optimization and Control This paper proposes LCFL, a novel clustering metric for evaluating clients' data distributions in federated learning. LCFL aligns with federated learning requirements, accurately assessing client-to-client variations in data distribution. It offers advantages over existing clustered federated learning methods, addressing privacy concerns, improving applicability to non-convex models, and providing more accurate classification results. LCFL does not require prior knowledge of clients' data distributions. We provide a rigorous mathematical analysis, demonstrating the correctness and feasibility of our framework. Numerical experiments with neural network instances highlight the superior performance of LCFL over baselines on several clustered federated learning benchmarks. |
| title | Novel clustered federated learning based on local loss |
| topic | Machine Learning Optimization and Control |
| url | https://arxiv.org/abs/2407.09360 |