BriefGPT.xyz
Dec, 2019
具有域自适应的私有联邦学习
Private Federated Learning with Domain Adaptation
HTML
PDF
Daniel Peterson, Pallika Kanani, Virendra J. Marathe
TL;DR
本文提出了一种基于联邦学习的机器学习分布式范例,可以在保护隐私的前提下进行多方联合重新训练共享模型,并通过用户级领域自适应来提高模型精度,实验结果表明在为FL模型强制实施差分隐私界限时,该技术可以更大程度地提升模型的准确性。
Abstract
federated learning
(FL) is a
distributed machine learning
(ML) paradigm that enables multiple parties to jointly re-train a shared model without sharing their data with any other parties, offering advantages in b
→