PulseAugur
中
实时 14:29:25
English(EN) When to spend privacy: optimal allocation in private learning

新方法优化机器学习训练中的隐私花费

研究人员开发了一种优化机器学习中隐私分配的新方法,特别针对差分隐私随机梯度下降(DPSGD)。该方法将高斯噪声方差调度表述为在集中差分隐私约束下的控制问题。这项工作引入了隐私预算伴随镜面下降(PB-AMD),它结合了伴随梯度和熵更新,以确保正值和预算遵守。该方法旨在通过优化迭代过程中的隐私花费来提高训练效用,其性能优于现有的分配规则,并且与投影梯度下降相比优化速度更快。 AI

影响 这项研究通过优化训练过程中隐私预算的使用方式,有望实现更高效、更有效的隐私保护机器学习模型。

排序理由 学术论文,详细介绍了机器学习中隐私分配的新方法。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

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

新方法优化机器学习训练中的隐私花费

本文如何被排名

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
学术论文,详细介绍了机器学习中隐私分配的新方法。[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
3 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. arXiv cs.AI TIER_1 English(EN) · Umberto Biccari, Yongcun Song ·

    何时投入隐私:私有学习中的最优分配

    arXiv:2610.39954v1 Announce Type: cross Abstract: Privacy accounting bounds the cumulative privacy cost of Differentially Private Stochastic Gradient Descent but does not determine how to allocate a fixed budget across iterations. We formulate the complete Gaussian noise-variance…