Cross-Training with Multi-View Knowledge Fusion for Heterogenous Federated Learning

2024-05-30

arXiv.org

AI Summary

To address local-to-global knowledge forgetting and feature-space inconsistency caused by data distribution shifts in heterogeneous federated learning, this paper proposes FedCT, a cross-training framework. Methodologically, FedCT introduces (1) consistency-aware knowledge broadcasting to robustly transfer global class prototypes to local models; (2) multi-view knowledge-guided representation learning that jointly models global class prototypes and local discriminative structures; and (3) a MixUp-enhanced feature alignment mechanism to improve cross-device generalization. Evaluated on four standard benchmarks, FedCT significantly outperforms state-of-the-art methods. Ablation studies and case analyses validate that the three components synergistically mitigate knowledge forgetting while enhancing representation stability and class discriminability.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Computer Vision: Representation Learning for Vision
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Federated learning benefits from cross-training strategies, which enables models to train on data from distinct sources to improve the generalization capability. However, the data heterogeneity between sources may lead models to gradually forget previously acquired knowledge when undergoing cross-training to adapt to new tasks or data sources. We argue that integrating personalized and global knowledge to gather information from multiple perspectives could potentially improve performance. To achieve this goal, this paper presents a novel approach that enhances federated learning through a cross-training scheme incorporating multi-view information. Specifically, the proposed method, termed FedCT, includes three main modules, where the consistency-aware knowledge broadcasting module aims to optimize model assignment strategies, which enhances collaborative advantages between clients and achieves an efficient federated learning process. The multi-view knowledge-guided representation learning module leverages fused prototypical knowledge from both global and local views to enhance the preservation of local knowledge before and after model exchange, as well as to ensure consistency between local and global knowledge. The mixup-based feature augmentation module aggregates rich information to further increase the diversity of feature spaces, which enables the model to better discriminate complex samples. Extensive experiments were conducted on four datasets in terms of performance comparison, ablation study, in-depth analysis and case study. The results demonstrated that FedCT alleviates knowledge forgetting from both local and global views, which enables it outperform state-of-the-art methods.

Problem

• Addresses data heterogeneity in federated learning cross-training
• Aligns local and global knowledge via multi-view distillation
• Enhances feature diversity to improve sample discrimination

Innovation

• Consistency-aware knowledge broadcasting for model assignment
• Multi-view knowledge-guided representation learning
• Mixup-based feature augmentation for diversity