EUVIMEDEUVIMED entdecken
Crossref

ClusterFL: A Clustering-based Federated Learning System for Human Activity Recognition

Xiaomin Ouyang, Zhiyuan Xie, Jiayu Zhou, Guoliang Xing, Jianwei Huang

ACM Transactions on Sensor Networks · 28.02.2023 · 19 (1) · 1-32

DOI: 10.1145/3554980 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Transactions on Sensor Networks, online erschienen am 08.12.2022. Verfasst von Xiaomin Ouyang und 4 weiteren Autor:innen. Die Arbeit wurde bisher 78-mal zitiert (Crossref) und nennt 47 Literaturangaben. Gefördert von: Research Grants Council (RGC) of Hong Kong, China, Alzheimer’s Drug Discovery Foundation, Shenzhen Science and Technology.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Federated Learning (FL) has recently received significant interest, thanks to its capability of protecting data privacy. However, existing FL paradigms yield unsatisfactory performance for a wide class of human activity recognition (HAR) applications, since they are oblivious to the intrinsic relationship between data of different users. We propose ClusterFL, a clustering-based federated learning system that can provide high model accuracy and low communication overhead for HAR applications. ClusterFL features a novel clustered multi-task federated learning framework that minimizes the empirical training loss of multiple learned models while automatically capturing the intrinsic clustering relationship among the nodes. We theoretically prove the convergence of proposed FL framework for non-convex and strongly convex models and provide the guidance on selection of hyper-parameters for achieving such convergence. Based on the learned cluster relationship, ClusterFL can efficiently drop the nodes that converge slower or have little correlations with others in each cluster, significantly speeding up the convergence while maintaining the accuracy performance. We evaluate the performance of ClusterFL on an NVIDIA edge testbed using four new HAR datasets collected from 145 users. The results show that ClusterFL outperforms several state-of-the-art FL paradigms in terms of overall accuracy and can save more than 50% communication overhead.

Förderung

  • Research Grants Council (RGC) of Hong Kong, China
  • Alzheimer’s Drug Discovery Foundation
  • Shenzhen Science and Technology
  • Guangdong Basic and Applied Basic Research Foundation
  • Shenzhen Institute of Artificial Intelligence and Robotics for Society

Zitieren

APA

Ouyang, X., Xie, Z., Zhou, J., Xing, G., & Huang, J. (2023). ClusterFL: A Clustering-based Federated Learning System for Human Activity Recognition. ACM Transactions on Sensor Networks, 19(1), 1-32. https://doi.org/10.1145/3554980

Vancouver

Ouyang X, Xie Z, Zhou J, Xing G, Huang J. ClusterFL: A Clustering-based Federated Learning System for Human Activity Recognition. ACM Transactions on Sensor Networks. 2023;19(1):1-32. doi:10.1145/3554980

BibTeX
@article{Ouyang2023,
  author = {Ouyang, Xiaomin and Xie, Zhiyuan and Zhou, Jiayu and Xing, Guoliang and Huang, Jianwei},
  title = {ClusterFL: A Clustering-based Federated Learning System for Human Activity Recognition},
  journal = {ACM Transactions on Sensor Networks},
  year = {2023},
  volume = {19},
  number = {1},
  pages = {1--32},
  doi = {10.1145/3554980}
}
Kostenlos starten →