Robust and IP-Protecting Vertical Federated Learning against Unexpected Quitting of Parties

2023/03/28 by Jingwei Sun, Zhixu Du, Sun, Jingwei +11 · 2 citations
Computer Science · Engineering · Medicine · #COVID-19 diagnosis using AI #Cryptography and Security (cs.CR) #FOS: Computer and information sciences #Geophysical Methods and Applications #Machine Learning (cs.LG) #Privacy-Preserving Technologies in Data

paper · pdf · doi:10.48550/arxiv.2303.18178

openalex publication_date 2023/03/28 · openalex created_date 2023/04/06 · openalex updated_date 2026/07/28

Abstract

Vertical federated learning (VFL) enables a service provider (i.e., active party) who owns labeled features to collaborate with passive parties who possess auxiliary features to improve model performance. Existing VFL approaches, however, have two major vulnerabilities when passive parties unexpectedly quit in the deployment phase of VFL - severe performance degradation and intellectual property (IP) leakage of the active party's labels. In this paper, we propose Party-wise Dropout to improve the VFL model's robustness against the unexpected exit of passive parties and a defense method called DIMIP to protect the active party's IP in the deployment phase. We evaluate our proposed methods on multiple datasets against different inference attacks. The results show that Party-wise Dropout effectively maintains model performance after the passive party quits, and DIMIP successfully disguises label information from the passive party's feature extractor, thereby mitigating IP leakage.

Cited by

• Reliable Vertical Federated Learning in 5G Core Network Architecture
• Deep Latent Variable Model based Vertical Federated Learning with Flexible Alignment and Labeling Scenarios

Related

• AntiFLipper: A Secure and Efficient Defense Against Label-Flipping Attacks in Federated Learning
• Just a Simple Transformation is Enough for Data Protection in Vertical Federated Learning
• REMISVFU: Vertical Federated Unlearning via Representation Misdirection for Intermediate Output Feature
• VFLIP: A Backdoor Defense for Vertical Federated Learning via Identification and Purification
• Mutual Information Regularization for Vertical Federated Learning