Privacy With Estimation Guarantees
Privacy With Estimation Guarantees
复制标题
DOI：
10.1109/tit.2019.2934414
复制
发表时间：
2017-10
影响因子：
2.5
通讯作者：
H. Wang;Lisa Vo;F. Calmon;M. Médard;K. Duffy;Mayank Varia
中科院分区：
文献类型：
--
作者：
H. Wang;Lisa Vo;F. Calmon;M. Médard;K. Duffy;Mayank Varia
我们研究数据隐私的核心问题：如何与分析师共享数据，同时向拥有数据的用户提供隐私和实用程序保证。在这种情况下，我们提供了对隐私 - 实用性权衡（PUT）的估计理论分析。在这里，允许分析师重建数据的某些功能（实用程序），而其他私人功能则不应以低于一定阈值（隐私）的失真重建。我们展示了卡方信息如何捕获本案中的基本投入，并为最佳投票提供了界限。我们提出了一个凸面程序，以计算披露和隐藏的功能的先验性并已知数据时，可以计算隐私映射。当使用经验分布来计算隐私映射而不是真实的数据分布时，我们会根据最小于点误差得出最小于点误差，并评估我们方法的鲁棒性。我们通过两个数值实验说明了所提出的方法。
We study the central problem in data privacy: how to share data with an analyst while providing both privacy and utility guarantees to the user that owns the data. In this setting, we present an estimation-theoretic analysis of the privacy-utility trade-off (PUT). Here, an analyst is allowed to reconstruct (in a mean-squared error sense) certain functions of the data (utility), while other private functions should not be reconstructed with distortion below a certain threshold (privacy). We demonstrate how chi-square information captures the fundamental PUT in this case and provide bounds for the best PUT. We propose a convex program to compute privacy-assuring mappings when the functions to be disclosed and hidden are known a priori and the data distribution is known. We derive lower bounds on the minimum mean-squared error of estimating a target function from the disclosed data and evaluate the robustness of our approach when an empirical distribution is used to compute the privacy-assuring mappings instead of the true data distribution. We illustrate the proposed approach through two numerical experiments.