信息安全研究2024，Vol.10Issue(6)：506-512,7.DOI:10.12379/j.issn.2096-1057.2024.06.03
一种安全高效的全匿踪纵向联邦学习方法
A Secure and Efficient Method of Fully Anonymous Vertical Federated Learning
尤志强 1李月 2姜玮 3方竞 1陈立峰 1卞阳1
作者信息
- 1. 上海富数科技有限公司 上海 200120
- 2. 中国电子口岸数据中心上海分中心 上海 200120
- 3. 上海海关科技处 上海 200135
- 折叠
摘要
Abstract
As a key technical paradigm to achieve"data availability and invisibility",the core process of vertical federated learning is sample alignment based on private set intersection.Although the private set intersection protects the privacy of non-intersected information,it can't meet the privacy protection requirements of user IDs in the intersected set.This paper proposes a fully anonymous vertical federated learning framework based on anonymous alignment to ensure that no private information of each holder set will be disclosed during the whole process.An implementation framework based on secure multi-party computation is proposed for fully anonymous joint modeling.The high performance and low error characteristics of the framework are verified through experiments,indicating it can be better applied in practice.关键词
纵向联邦学习/安全求交/匿踪求交/多方安全计算/匿踪学习Key words
vertical federated learning/private set intersection/anonymous alignment/secure multi-party computation/anonymous federated learning分类
信息技术与安全科学引用本文复制引用
尤志强,李月,姜玮,方竞,陈立峰,卞阳..一种安全高效的全匿踪纵向联邦学习方法[J].信息安全研究,2024,10(6):506-512,7.
