Federated learning is a machine learning paradigm where many clients cooperatively train a single centralized model while keeping their data private and decentralized. This novel paradigm imposes many challenges, such as ...