arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2406.02599cs.CRcs.AI
收藏

基于线性规划的隐私感知随机量化

Privacy-Aware Randomized Quantization via Linear Programming

Ohio State University(俄亥俄州立大学);Yahoo Research(雅虎研究院)

Zhongteng Cai, Xueru Zhang, Mohammad Mahdi Khalili

AI总结:

本文针对差分隐私下离散输出存在偏差或权衡不佳的问题,提出一类无偏且具高度自由度的量化机制,并通过线性规划求解最优配置,实现更优的隐私-准确率权衡。

摘要:

差分隐私机制(如Gaussian或Laplace机制)已广泛用于数据分析中以保护个人隐私。然而,它们大多为连续输出而设计,不适用于需要离散值的场景。尽管最近提出了各种量化机制以在差分隐私下生成离散输出,但其结果要么存在偏差,要么在准确率与隐私的权衡上表现较差。本文提出了一类无偏且满足差分隐私的量化机制。该机制具有高度自由度,且现有某些机制可视为其特例。为寻找最优机制,我们构建了一个线性优化问题,可借助线性规划工具高效求解。实验表明,与基线方法相比,所提机制能实现更优的隐私与准确率权衡。

英文摘要:

Differential privacy mechanisms such as the Gaussian or Laplace mechanism have been widely used in data analytics for preserving individual privacy. However, they are mostly designed for continuous outputs and are unsuitable for scenarios where discrete values are necessary. Although various quantization mechanisms were proposed recently to generate discrete outputs under differential privacy, the outcomes are either biased or have an inferior accuracy-privacy trade-off. In this paper, we propose a family of quantization mechanisms that is unbiased and differentially private. It has a high degree of freedom and we show that some existing mechanisms can be considered as special cases of ours. To find the optimal mechanism, we formulate a linear optimization that can be solved efficiently using linear programming tools. Experiments show that our proposed mechanism can attain a better privacy-accuracy trade-off compared to baselines.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑