Skip to main navigation Skip to search Skip to main content

A Reinforcement Learning-Based Client Selection Strategy for Efficient Federated Learning

  • Mingzhi Jiang
  • , Ziqi Song
  • , Chen Cui*
  • , Shen Wang
  • *Corresponding author for this work
  • Heilongjiang University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

With growing concerns over data privacy, federated learning (FL) enables collaborative model training without exposing raw data. However, FL faces persistent challenges in efficient client selection and effective resource utilization, especially under heterogeneous and Non-IID conditions. This paper presents a reinforcement learning-based client selection strategy leveraging a Dueling Double Deep Q-Network (DD-DQN) to enable adaptive and informed decision-making. Additionally, a novel aggregation mechanism is proposed, which assigns dynamic weights based on normalized Q-values, local training loss, and client data volume to improve aggregation accuracy. Experiments conducted on CIFAR-10, CIFAR-100, and MNIST under Non-IID partitions demonstrate that the proposed method improves global model accuracy by approximately 1.0% on average compared to FedAvg and FLASH-RL, while also achieving more stable convergence and higher system efficiency. These results validate the effectiveness of the proposed approach in enhancing federated learning performance under challenging data distributions.

Original languageEnglish
Title of host publicationAdvances in Intelligent Data Analysis and Applications - Proceedings of the 9th Euro-China Conference on Intelligent Data Analysis and Applications
EditorsVáclav Snášel, Lingping Kong, Chu-Chuan Chu, Jeng-Shyang Pan, Anna Esposito
PublisherSpringer Science and Business Media Deutschland GmbH
Pages202-215
Number of pages14
ISBN (Print)9783032130471
DOIs
StatePublished - 2026
Event9th Euro-China Conference on Intelligent Data Analysis and Applications, ECC 2025 - Ostrava, Czech Republic
Duration: 21 Jul 202523 Jul 2025

Publication series

NameSmart Innovation, Systems and Technologies
Volume466 SIST
ISSN (Print)2190-3018
ISSN (Electronic)2190-3026

Conference

Conference9th Euro-China Conference on Intelligent Data Analysis and Applications, ECC 2025
Country/TerritoryCzech Republic
CityOstrava
Period21/07/2523/07/25

Keywords

  • Client Selection
  • Federated Learning
  • Non-IID Data
  • Reinforcement Learning

Fingerprint

Dive into the research topics of 'A Reinforcement Learning-Based Client Selection Strategy for Efficient Federated Learning'. Together they form a unique fingerprint.

Cite this