PulseAugur
中
实时 11:09:55
English(EN) Counting on Dummies: Privacy Amplification with Observable Counts and Truncated Batches

新的DP-SGD方法通过隐藏批次大小来增强隐私

作者 PulseAugur 编辑部 · [1 个来源] ·

研究人员开发了一种新机制，通过对批次大小进行隐私化来增强差分隐私随机梯度下降（DP-SGD）中的隐私性。该方法使用随机数量的虚拟点来隐藏实际批次大小，即使在批次因计算限制而被截断的情况下也能维持隐私保证。该方法易于在分布式环境中实现，并与现有方法相比提供了更好的隐私-效率权衡，尤其适用于泊松采样。 AI

影响 增强了使用DP-SGD训练的机器学习模型的隐私保证，有可能增加其在敏感数据应用中的采用。

排序理由 关于DP-SGD中隐私放大新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

新的DP-SGD方法通过隐藏批次大小来增强隐私

本文如何被排名

Signal score
1 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
关于DP-SGD中隐私放大新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, safety
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
1 days old
Coverage has settled into its steady-state source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. arXiv cs.LG TIER_1 English(EN) · Vitaly Feldman, Audra McMillan, Kunal Talwar ·

    Counting on Dummies: Privacy Amplification with Observable Counts and Truncated Batches

    arXiv:2610.42139v1 Announce Type: new Abstract: In many practical deployments of Differentially Private Stochastic Gradient Descent (DP-SGD), the server processes data in batches. Randomly sampling these batches allows for better privacy bounds via privacy amplification results. …

相关实体

相关话题