Noun
federated learning (uncountable) (artificial intelligence) A machine learning technique that trains an algorithm across multiple decentralized edge devices or servers holding local data samples, without exchanging them.