DFedReweighting: A Unified Framework for Objective-Oriented Reweighting in Decentralized Federated Learning
This work addresses fairness and robustness issues for decentralized federated learning systems, representing an incremental advancement in aggregation methods.
The paper tackles challenges in decentralized federated learning, such as fairness and robustness, by proposing DFedReweighting, a unified aggregation framework that uses objective-oriented reweighting, and it demonstrates linear convergence and significant improvements in fairness and robustness against Byzantine attacks in experiments.
Decentralized federated learning (DFL) has recently emerged as a promising paradigm that enables multiple clients to collaboratively train machine learning model through iterative rounds of local training, communication, and aggregation without relying on a central server which introduces potential vulnerabilities in conventional Federated Learning. Nevertheless, DFL systems continue to face a range of challenges, including fairness, robustness, etc. To address these challenges, we propose \textbf{DFedReweighting}, a unified aggregation framework designed to achieve diverse objectives in DFL systems via a objective-oriented reweighting aggregation at the final step of each learning round. Specifically, the framework first computes preliminary weights based on \textit{target performance metric} obtained from auxiliary dataset constructed using local data. These weights are then refined using \textit{customized reweighting strategy}, resulting in the final aggregation weights. Our results from the theoretical analysis demonstrate that the appropriate combination of the target performance metric and the customized reweighting strategy ensures linear convergence. Experimental results consistently show that our proposed framework significantly improves fairness and robustness against Byzantine attacks in diverse scenarios. Provided that appropriate target performance metrics and customized reweighting strategy are selected, our framework can achieve a wide range of desired learning objectives.