Federated Machine Learning

TLDR

Today’s artificial intelligence still faces two major challenges: data isolation across industries and growing data privacy and security concerns. The study proposes secure federated learning to address these challenges. They present a comprehensive secure federated‑learning framework comprising horizontal, vertical, and transfer learning, detailing its architecture, applications, and a survey of related work, and suggest using federated mechanisms to build inter‑organizational data networks.

Abstract

Today’s artificial intelligence still faces two major challenges. One is that, in most industries, data exists in the form of isolated islands. The other is the strengthening of data privacy and security. We propose a possible solution to these challenges: secure federated learning. Beyond the federated-learning framework first proposed by Google in 2016, we introduce a comprehensive secure federated-learning framework, which includes horizontal federated learning, vertical federated learning, and federated transfer learning. We provide definitions, architectures, and applications for the federated-learning framework, and provide a comprehensive survey of existing works on this subject. In addition, we propose building data networks among organizations based on federated mechanisms as an effective solution to allowing knowledge to be shared without compromising user privacy.