FedPBS: Proximal-Balanced Scaling Federated Learning Model for Robust Personalized Training for Non-IID Data

Researchers propose FedPBS, a new federated learning algorithm that addresses key challenges in distributed AI training including statistical heterogeneity and uneven client participation. The algorithm dynamically adapts batch sizes and applies proximal corrections to improve model convergence while preserving data privacy across distributed clients.

• →FedPBS combines techniques from FedBS and FedProx to create a more robust federated learning approach for non-IID data scenarios.
• →The algorithm dynamically adapts batch sizes to client resources and applies proximal corrections to stabilize training.
• →Experiments show consistent outperformance over state-of-the-art methods including FedBS, FedGA, MOON, and FedProx on benchmark datasets.
• →The approach demonstrates stable convergence under extreme data heterogeneity conditions in federated environments.
• →Applications span healthcare, finance, mobility, and smart-city systems where data privacy is critical.