FedDyH: A Multi-Policy with GA Optimization Framework for Dynamic Heterogeneous Federated Learning
Open Access
- 17 March 2025
- journal article
- research article
- Published by MDPI AG in Biomimetics
- Vol. 10 (3) , 185
Abstract
Federated learning (FL) is a distributed learning technique that ensures data privacy and has shown significant potential in cross-institutional image analysis. However, existing methods struggle with the inherent dynamic heterogeneity of real-world data, such as changes in cellular differentiation during disease progression or feature distribution shifts due to different imaging devices. This dynamic heterogeneity can cause catastrophic forgetting, leading to reduced performance in medical predictions across stages. Unlike previous federated learning studies that paid insufficient attention to dynamic heterogeneity, this paper proposes the FedDyH framework to address this challenge. Inspired by the adaptive regulation mechanisms of biological systems, this framework incorporates several core modules to tackle the issues arising from dynamic heterogeneity. First, the framework simulates intercellular information transfer through cross-client knowledge distillation, preserving local features while mitigating knowledge forgetting. Additionally, a dynamic regularization term is designed in which the strength can be adaptively adjusted based on real-world conditions. This mechanism resembles the role of regulatory T cells in the immune system, balancing global model convergence with local specificity adjustments to enhance the robustness of the global model while preventing interference from diverse client features. Finally, the framework introduces a genetic algorithm (GA) to simulate biological evolution, leveraging mechanisms such as gene selection, crossover, and mutation to optimize hyperparameter configurations. This enables the model to adaptively find the optimal hyperparameters in an ever-changing environment, thereby improving both adaptability and performance. Prior to this work, few studies have explored the use of optimization algorithms for hyperparameter tuning in federated learning. Experimental results demonstrate that the FedDyH framework improves accuracy compared to the SOTA baseline FedDecorr by 2.59%, 0.55%, and 5.79% on the MNIST, Fashion-MNIST, and CIFAR-10 benchmark datasets, respectively. This framework effectively addresses data heterogeneity issues in dynamic heterogeneous environments, providing an innovative solution for achieving more stable and accurate distributed federated learning.Keywords
Funding Information
- National Natural Science Foundation of China (61272315, 82303675, LY21F020028, LY24F030005, 2023C01040, 19YJAZH047, GJJ218605)
- Natural Science Foundation of Zhejiang Province (61272315, 82303675, LY21F020028, LY24F030005, 2023C01040, 19YJAZH047, GJJ218605)
- Science and Technology Project of Zhejiang Province (61272315, 82303675, LY21F020028, LY24F030005, 2023C01040, 19YJAZH047, GJJ218605)
- Ministry of Education of Humanities and Social Science Project (61272315, 82303675, LY21F020028, LY24F030005, 2023C01040, 19YJAZH047, GJJ218605)
- Jiangxi Provincial Department of Education Science and Technology Research Project (61272315, 82303675, LY21F020028, LY24F030005, 2023C01040, 19YJAZH047, GJJ218605)
This publication has 17 references indexed in Scilit:
- Enhancing sports image data classification in federated learning through genetic algorithm-based optimization of base architecturePLOS ONE, 2024
- Class-incremental learning with Balanced Embedding Discrimination MaximizationNeural Networks, 2024
- GA Approach to Optimize Training Client Set in Federated LearningIEEE Access, 2023
- Manta ray foraging optimization based on mechanics game and progressive learning for multiple optimization problemsApplied Soft Computing, 2023
- Experimental Evaluation and Analysis of Federated Learning in Edge Computing EnvironmentsIEEE Access, 2023
- A contemplative perspective on federated machine learning: Taxonomy, threats & vulnerability assessment and challengesJournal of King Saud University Computer and Information Sciences, 2022
- Adaptive Client Clustering for Efficient Federated Learning Over Non-IID and Imbalanced DataIEEE Transactions on Big Data, 2022
- Federated Learning for Internet of Things: A Comprehensive SurveyIEEE Communications Surveys & Tutorials, 2021
- Learning Deep Representations with Probabilistic Knowledge Transfer2018
- The MNIST Database of Handwritten Digit Images for Machine Learning Research [Best of the Web]IEEE Signal Processing Magazine, 2012