Incentive Mechanism for Maximizing Social Surplus in Cross-Device Federated Learning

DOI

説明

In cross-device federated learning, mobile devices collabo<br />rate on machine learning via a central server without sharing<br /> data. In this paper, we propose an incentive mechanism for<br /> maximizing social surplus. In this mechanism, a distributed<br /> algorithm is utilized in each mobile device.<br /> We proposed an incentive mechanism for maximizing so<br />cial surplus. From numerical examples, we found that our proposed method can maximize the social surplus without sharing any information.

収録刊行物

  • IEICE Proceeding Series

    IEICE Proceeding Series 85 P2-13-, 2024-11-25

    The Institute of Electronics, Information and Communication Engineers

キーワード

詳細情報 詳細情報について

  • CRID
    1390584088487551872
  • DOI
    10.34385/proc.85.p2-13
  • ISSN
    21885079
  • 本文言語コード
    en
  • データソース種別
    • JaLC
  • 抄録ライセンスフラグ
    使用不可

問題の指摘

ページトップへ