Adaptive differential privacy boosts asynchronous federated learning accuracy across simulated satellite networks
Computer scientists have developed an asynchronous federated learning framework that coordinates client trust, adaptive noise scaling, and delay weighting to ...

