Differentially private distributed logistic regression using private and public data

By Zhanglong Ji, Xiaoqian Jiang, Shuang Wang, Li Xiong, Lucila Ohno-Machado. 2014

58 citations recorded by the source corpus.

Privacy protecting is an important issue in medical informatics and differential privacy is a state-of-the-art framework for data privacy research. Differential privacy offers provable privacy against attackers who have auxiliary information, and can be applied to data mining models (for example, logistic regression). However, differentially private methods sometimes introduce too much noise and make outputs less useful. Given available public data in medical research (e.g. from patients who sign open-consent agreements), we can design algorithms that use both public and private data sets to decrease the amount of noise that is introduced. In this paper, we modify the update step in Newton-Raphson method to propose a differentially private distributed logistic regression model based on both public and private data. We try our algorithm on three different data sets, and show its advantage over: (1) a logistic regression model based solely on public data, and (2) a differentially private distributed logistic regression model based on private data under various scenarios. Logistic regression models built with our new algorithm based on both private and public datasets demonstrate better utility than models that trained on private or public datasets alone without sacrificing the rigorous privacy guarantee.

Subfields: Artificial Intelligence, Statistics and Probability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistics and Probability

Citations in the corpus

Learning privately from multiparty data (2016, background): “Besides, efficient differentially private mechanisms are known only for certain types of classifiers so far (see (Ji et al., 2014) for a review.)”
Differential Privacy and Machine Learning: a Survey and Review (2014, background): “According to [28, 29], such a public dataset can enhance the performance of differentially private mechanisms.”
How to DP-fy ML: A Practical Guide to Machine Learning with Differential Privacy (2023, methodology): “is to use the public data to estimate the Hessian (while using private data to compute the noised gradients) (Ji et al., 2014).”
InPrivate Digging: Enabling Tree-based Distributed Data Mining with Differential Privacy (2018, methodology): “It has been widely used to design privacy-preserving schemes for data publishing [18] and data mining tasks such as linear regression [6], support vector machines [7], and neural networks [9].”