A Fairness-Aware Peer-to-Peer Decentralized Learning Framework with Heterogeneous Devices

被引:6
|
作者
Chen, Zheyi [1 ]
Liao, Weixian [1 ]
Tian, Pu [1 ]
Wang, Qianlong [1 ]
Yu, Wei [1 ]
机构
[1] Towson Univ, Dept Comp & Informat Sci, Towson, MD 21252 USA
关键词
decentralized learning; learning fairness; heterogeneity; INTERNET;
D O I
10.3390/fi14050138
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Distributed machine learning paradigms have benefited from the concurrent advancement of deep learning and the Internet of Things (IoT), among which federated learning is one of the most promising frameworks, where a central server collaborates with local learners to train a global model. The inherent heterogeneity of IoT devices, i.e., non-independent and identically distributed (non-i.i.d.) data, and the inconsistent communication network environment results in the bottleneck of a degraded learning performance and slow convergence. Moreover, most weight averaging-based model aggregation schemes raise learning fairness concerns. In this paper, we propose a peer-to-peer decentralized learning framework to tackle the above issues. Particularly, each local client iteratively finds a learning pair to exchange the local learning model. By doing this, multiple learning objectives are optimized to advocate for learning fairness while avoiding small-group domination. The proposed fairness-aware approach allows local clients to adaptively aggregate the received model based on the local learning performance. The experimental results demonstrate that the proposed approach is capable of significantly improving the efficacy of federated learning and outperforms the state-of-the-art schemes under real-world scenarios, including balanced-i.i.d., unbalanced-i.i.d., balanced-non.i.i.d., and unbalanced-non.i.i.d. environments.
引用
收藏
页数:18
相关论文
共 50 条
  • [1] Framework for peer-to-peer distributed computing in a heterogeneous decentralized environment
    Verbeke, J
    Nadgir, N
    Ruetsch, G
    Sharapov, I
    GRID COMPUTING - GRID 2002, 2002, 2536 : 1 - 12
  • [2] On the (In)security of Peer-to-Peer Decentralized Machine Learning
    Pasquini, Dario
    Raynal, Mathilde
    Troncoso, Carmela
    2023 IEEE SYMPOSIUM ON SECURITY AND PRIVACY, SP, 2023, : 418 - 436
  • [3] DeFTA: A plug-and-play peer-to-peer decentralized federated learning framework
    Zhou, Yuhao
    Shi, Minjia
    Tian, Yuxin
    Ye, Qing
    Lv, Jiancheng
    INFORMATION SCIENCES, 2024, 670
  • [4] Skin Lesion Classification using a Decentralized Peer-to-Peer Federated Learning Framework
    Sarkar, Sujan
    Majumder, Siddhant
    Mandal, Diptarka
    Gulvanskii, Vyacheslav
    Kaplun, Dmitrii
    Sarkar, Ram
    2024 20TH INTERNATIONAL SYMPOSIUM ON MEDICAL INFORMATION PROCESSING AND ANALYSIS, SIPAIM 2024, 2024,
  • [5] Peer-to-Peer Secure Updates for Heterogeneous Edge Devices
    Herry, Herry
    Band, Emily
    Perkins, Colin
    Singer, Jeremy
    NOMS 2018 - 2018 IEEE/IFIP NETWORK OPERATIONS AND MANAGEMENT SYMPOSIUM, 2018,
  • [6] Peer-to-peer decentralized energy trading framework for retailers and prosumers
    Mehdinejad, Mehdi
    Shayanfar, Heidarali
    Mohammadi-Ivatloo, Behnam
    APPLIED ENERGY, 2022, 308
  • [7] Decentralized Peer-to-Peer Auctions
    Marcus Fontoura
    Mihail Ionescu
    Naftaly Minsky
    Electronic Commerce Research, 2005, 5 (1) : 7 - 24
  • [8] Fairness issues in peer-to-peer networks
    Wong, WS
    PROCEEDINGS OF THE 41ST IEEE CONFERENCE ON DECISION AND CONTROL, VOLS 1-4, 2002, : 2005 - 2010
  • [9] Peer-to-peer auctioning with fairness incentives
    Rozanski, Bartlomiej
    Proceedings of the Fifth IASTED International Conference on Communication Systems and Networks, 2006, : 177 - 182
  • [10] Proportional fairness in heterogeneous peer-to-peer networks through reciprocity and Gibbs sampling
    Zubeldia, Martin
    Ferragut, Andres
    Paganini, Fernando
    2013 51ST ANNUAL ALLERTON CONFERENCE ON COMMUNICATION, CONTROL, AND COMPUTING (ALLERTON), 2013, : 123 - 130