FedDAG: Clustered Federated Learning via Global Data and Gradient Integration for Heterogeneous Environments

Researchers introduce FedDAG, a new clustered federated learning framework that improves AI model training across heterogeneous client environments. The system combines data and gradient similarity metrics for better client clustering and uses a dual-encoder architecture to enable knowledge sharing across clusters while maintaining specialization.

• →FedDAG addresses performance drops in federated learning when client data varies significantly across participants.
• →The framework uses a weighted, class-wise similarity metric combining both data and gradient information for more accurate client clustering.
• →A dual-encoder architecture enables cross-cluster feature transfer while preserving cluster-specific model specialization.
• →Experiments show FedDAG consistently outperforms existing clustered federated learning approaches in accuracy.
• →The research advances distributed AI training capabilities for heterogeneous environments without compromising data privacy.