English(EN) On the Stability of Spherical Hellinger-Kantorovich Flows and Their Implications for Differential Privacy

新理论为采样提供差分隐私保证

作者 PulseAugur 编辑部 · [2 个来源] · 2026-05-22 17:38

研究人员开发了球形 Hellinger-Kantorovich (SHK) 梯度流的扰动理论，能够精确比较基于不同势能的流。该理论为对数似然比和散度提供了无量纲界限，并将其应用于差分隐私机制的近似采样。研究结果为基于 SHK 的采样器和近似差分隐私证书提供了明确的 Pure-DP 保证。 AI

影响这项研究为机器学习中的差分隐私提供了新的理论工具，有可能提高 AI 模型所用数据的安全性。

排序理由这是一篇发表在 arXiv 上的研究论文，详细介绍了一个新的理论框架及其应用。

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv stat.ML TIER_1 English(EN) · Aratrika Mustafi, Soumya Mukherjee · 2026-05-25 04:00

On the Stability of Spherical Hellinger-Kantorovich Flows and Their Implications for Differential Privacy

arXiv:2605.23879v1 Announce Type: new Abstract: Gradient-flow sampling interprets a Gibbs distribution as the minimizer of an energy functional over probability measures and generates dynamics converging to this target. Under spherical Hellinger-Kantorovich (SHK) geometry, the fl…
arXiv stat.ML TIER_1 English(EN) · Soumya Mukherjee · 2026-05-22 17:38

On the Stability of Spherical Hellinger-Kantorovich Flows and Their Implications for Differential Privacy

Gradient-flow sampling interprets a Gibbs distribution as the minimizer of an energy functional over probability measures and generates dynamics converging to this target. Under spherical Hellinger-Kantorovich (SHK) geometry, the flow couples transport and reaction and coincides …