EUVIMEDEUVIMED entdecken
Crossref

Adaptive Federated Learning With Non-IID Data

Yan Zeng, Yuankai Mu, Junfeng Yuan, Siyuan Teng, Jilin Zhang, Jian Wan, Yongjian Ren, Yunquan Zhang

The Computer Journal · 14.11.2023 · 66 (11) · 2758-2772

DOI: 10.1093/comjnl/bxac118 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The Computer Journal, online erschienen am 30.09.2022. Verfasst von Yan Zeng und 7 weiteren Autor:innen. Die Arbeit wurde bisher 47-mal zitiert (Crossref) und nennt 36 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the widespread use of Internet of things(IoT) devices, it generates an enormous volume of data, and it is a challenge to mine the IoT data value while ensuring security and privacy. Federated learning is a decentralized approach for training data located on edge devices, such as mobile phones and IoT devices, while keeping privacy, efficiency, and security. However, the Non-IID (non-independent and identically distributed) data, always greatly impacts the performance of the global model. In this paper, we propose a FedDynamic algorithm to solve the statistical challenge of federated learning caused by Non-IID. As Non-IID data can lead to significant differences in model parameters between edge devices, we set different weights for different devices during model aggregation to get a high-performance global model. We analyze and exact key indices (local model accuracy, local data quality, and model difference between local models and the global model), which can reflect the quality of the model, and calculate the aggregation weight for edge devices based on the key indices. Furthermore, we dynamically adjust aggregation weight based on accuracy’s variety to solve weight staleness during the training process. Experiments on the MNIST, FMNIST, EMNIST, CINIC-10 and CIFAR-10 datasets show that the FedDynamic algorithm has better accuracy and convergence performance, compared to the FedAvg, FedProx and Scaffold algorithms.

Zitieren

APA

Zeng, Y., Mu, Y., Yuan, J., Teng, S., Zhang, J., Wan, J., Ren, Y., & Zhang, Y. (2023). Adaptive Federated Learning With Non-IID Data. The Computer Journal, 66(11), 2758-2772. https://doi.org/10.1093/comjnl/bxac118

Vancouver

Zeng Y, Mu Y, Yuan J, Teng S, Zhang J, Wan J, et al. Adaptive Federated Learning With Non-IID Data. The Computer Journal. 2023;66(11):2758-2772. doi:10.1093/comjnl/bxac118

BibTeX
@article{Zeng2023,
  author = {Zeng, Yan and Mu, Yuankai and Yuan, Junfeng and Teng, Siyuan and Zhang, Jilin and Wan, Jian and Ren, Yongjian and Zhang, Yunquan},
  title = {Adaptive Federated Learning With Non-IID Data},
  journal = {The Computer Journal},
  year = {2023},
  volume = {66},
  number = {11},
  pages = {2758--2772},
  doi = {10.1093/comjnl/bxac118}
}
Kostenlos starten →