{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/fed-rac-resource-aware-clustering-for","title":"Fed-RAC: Resource-Aware Clustering for Tackling Heterogeneity of Participants in Federated Learning","arxiv_id":null,"date":"2024-03-20","proceeding":"IEEE Transactions on Parallel and Distributed Systems 2024 3","authors":["Rahul Mishra","Hari Prabhat Gupta","Garvit Banga","Sajal K. Das"],"abstract":"Federated Learning is a training framework that\r\nenables multiple participants to collaboratively train a shared\r\nmodel while preserving data privacy. The heterogeneity of devices\r\nand networking resources of the participants delay the training\r\nand aggregation. The paper introduces a novel approach to federated learning by incorporating resource-aware clustering. This\r\nmethod addresses the challenges posed by the diverse devices and\r\nnetworking resources among participants. Unlike static clustering\r\napproaches, this paper proposes a dynamic method to determine\r\nthe optimal number of clusters using Dunn Indices. It enables\r\nadaptability to the varying heterogeneity levels among participants,\r\nensuring a responsive and customized approach to clustering. Next,\r\nthe paper goes beyond empirical observations by providing a mathematical derivation of the communication rounds for convergence\r\nwithin each cluster. Further, the participant assignment mechanism\r\nadds a layer of sophistication and ensures that devices and networking resources are allocated optimally. Afterwards, we incorporate\r\na leader-follower technique, particularly through knowledge distillation, which improves the performance of lightweight models\r\nwithin clusters. Finally, experiments are conducted to validate\r\nthe approach and to compare it with state-of-the-art. The results\r\ndemonstrated an accuracy improvement of over 3% compared to\r\nits closest competitor and a reduction in communication rounds of\r\naround 10%.","url_abs":"https://ieeexplore.ieee.org/document/10476717","url_pdf":"https://ieeexplore.ieee.org/document/10476717","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"fed-rac-resource-aware-clustering-for","repo_url":"https://github.com/GarvitBanga/Fed-RAC","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[{"task_slug":"clustering","task_name":"Clustering"},{"task_slug":"federated-learning","task_name":"Federated Learning"},{"task_slug":"knowledge-distillation","task_name":"Knowledge Distillation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}