Adapting Federated Learning to Dynamic and Non-Stationary Data: An Online Learning Approach

  • Kim, Jaemin; 
  • Hur, Donghyeon; 
  • Won, Dongwook; 
  • Cho, Sungrae
Citations

SCOPUS

0

초록

Online learning-based optimization represents a novel approach that integrates the principles of federated learning and online learning to effectively handle dynamic data distributions and continuous learning scenarios. This paper offers a comprehensive survey of the latest developments in online federated learning, focusing on optimizing communication efficiency in federated learning environments through online learning techniques. We delve into foundational concepts, key methodologies, challenges, and diverse applications of this emerg-ina field.

키워드

Federated Learning; Online Federated learning; Online Learning
제목
Adapting Federated Learning to Dynamic and Non-Stationary Data: An Online Learning Approach
저자
Kim, Jaemin; Hur, Donghyeon; Won, Dongwook; Cho, Sungrae
DOI
10.1109/ICAIIC64266.2025.10920733
발행일
2025-02
유형
Conference Paper
저널명
International Conference on Artificial Intelligence in Information and Communication, ICAIIC
페이지
12 ~ 14