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 2404.15409cs.LGcs.CRstat.ML
收藏

不足统计量扰动:用于私有最小二乘的稳定估计器

Insufficient Statistics Perturbation: Stable Estimators for Private Least Squares

University of Washington(华盛顿大学);Boston University(波士顿大学);Massachusetts Institute of Technology(麻省理工学院);Google Research(谷歌研究院);Harvard University(哈佛大学)

Gavin Brown, Jonathan Hayase, Samuel Hopkins, Weihao Kong, Xiyang Liu, Sewoong Oh, Juan C. Perdomo, Adam Smith

AI总结:

提出一种样本和时间高效的差分隐私普通最小二乘算法,其误差线性依赖于维度且独立于条件数,突破了先前算法在样本量、误差或时间上的限制。

摘要:

我们提出了一种样本和时间高效的差分隐私算法,用于普通最小二乘,其误差与维度线性相关,且独立于 $X^\ op X$ 的条件数,其中 $X$ 是设计矩阵。此前所有用于此任务的隐私算法均需要 $d^{3/2}$ 个样本、随条件数多项式增长的误差,或指数级时间。我们的近最优精度保证适用于任何具有有界统计杠杆率和有界残差的数据集。在技术上,我们基于 Brown 等人 (2023) 用于私有均值估计的方法,向经验回归向量的精心设计的稳定非私有估计器添加了缩放噪声。

英文摘要:

We present a sample- and time-efficient differentially private algorithm for ordinary least squares, with error that depends linearly on the dimension and is independent of the condition number of $X^\top X$, where $X$ is the design matrix. All prior private algorithms for this task require either $d^{3/2}$ examples, error growing polynomially with the condition number, or exponential time. Our near-optimal accuracy guarantee holds for any dataset with bounded statistical leverage and bounded residuals. Technically, we build on the approach of Brown et al. (2023) for private mean estimation, adding scaled noise to a carefully designed stable nonprivate estimator of the empirical regression vector.

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。

↑