Instruction: Outline a Federated Learning system employing homomorphic encryption to ensure data privacy during aggregation.
Context: Candidates must demonstrate their knowledge of homomorphic encryption and its application in designing privacy-preserving Federated Learning systems.
Official answer available
Preview the opening of the answer, then unlock the full walkthrough.
I would use homomorphic encryption when the privacy requirement justifies the added compute and communication cost. The general design is local training on clients, encrypted update generation, server-side aggregation in encrypted form where feasible, and decryption only at the stage...