Vertical Federated Learning: Concepts, Advances and Challenges

TL;DR

Vertical Federated Learning enables multiple parties with different features of the same users to collaboratively train models while preserving data privacy, with recent advances addressing effectiveness, efficiency, and privacy challenges.

Contribution

The paper provides a comprehensive review of VFL concepts, algorithms, privacy protocols, and introduces a unified framework VFLow considering multiple constraints.

Findings

Categorization of VFL settings and privacy protocols

Analysis of privacy attacks and defenses

Proposal of the VFLow framework for holistic VFL problem solving

Abstract

Vertical Federated Learning (VFL) is a federated learning setting where multiple parties with different features about the same set of users jointly train machine learning models without exposing their raw data or model parameters. Motivated by the rapid growth in VFL research and real-world applications, we provide a comprehensive review of the concept and algorithms of VFL, as well as current advances and challenges in various aspects, including effectiveness, efficiency, and privacy. We provide an exhaustive categorization for VFL settings and privacy-preserving protocols and comprehensively analyze the privacy attacks and defense strategies for each protocol. In the end, we propose a unified framework, termed VFLow, which considers the VFL problem under communication, computation, privacy, as well as effectiveness and fairness constraints. Finally, we review the most recent advances in industrial applications, highlighting open challenges and future directions for VFL.