Abstract:Traditional approaches to differential privacy assume a fixed privacy requirement ε for a computation, and attempt to maximize the accuracy of the computation subject to the privacy constraint. As differential privacy is increasingly deployed in practical settings, it may often be that there is instead a fixed accuracy requirement for a given computation and the data analyst would like to maximize the privacy of the computation subject to the accuracy constraint. This raises the question of how to find and run a maximally private empirical risk minimizer subject to a given accuracy requirement. We propose a general “noise reduction” framework that can apply to a variety of private empirical risk minimization (ERM) algorithms, using them to “search” the space of privacy levels to find the empirically strongest one that meets the accuracy constraint, and incurring only logarithmic overhead in the number of privacy levels searched. The privacy analysis of our algorithm leads naturally to a version of differential privacy where the privacy parameters are dependent on the data, which we term ex-post privacy, and which is related to the recently introduced notion of privacy odometers. We also give an ex-post privacy analysis of the classical AboveThreshold privacy tool, modifying it to allow for queries chosen depending on the database. Finally, we apply our approach to two common objective functions, regularized linear and logistic regression, and empirically compare our noise reduction methods to (i) inverting the theoretical utility guarantees of standard private ERM algorithms and (ii) a stronger, empirical baseline based on binary search.

Functional Mechanism: Regression Analysis under Differential Privacy

Differentially Private Log-Location-Scale Regression Using Functional Mechanism

The optimal mechanism in differential privacy

Differential Privacy: An Economic Method for Choosing Epsilon

Striking a Balance: An Optimal Mechanism Design for Heterogenous Differentially Private Data Acquisition for Logistic Regression

Differentially Private and Fair Classification Via Calibrated Functional Mechanism.

The Cost of Privacy: Optimal Rates of Convergence for Parameter Estimation with Differential Privacy

Differentially private regression analysis with dynamic privacy allocation

Instance-optimal Mean Estimation Under Differential Privacy

A Statistical Framework for Differential Privacy

Differentially Private Simple Linear Regression

Accuracy First: Selecting a Differential Privacy Level for Accuracy-Constrained ERM

Improving the Privacy and Practicality of Objective Perturbation for Differentially Private Linear Learners

Evaluating Differentially Private Machine Learning in Practice

Differentially Private Empirical Risk Minimization

Efficient Sparse Least Absolute Deviation Regression with Differential Privacy

Differential Privacy via Distributionally Robust Optimization

Minimax Rates of Estimating Approximate Differential Privacy

Information Design for Differential Privacy

Not All Attributes are Created Equal: $d_{\mathcal{X}}$-Private Mechanisms for Linear Queries

Differential Private Discrete Noise Adding Mechanism: Conditions, Properties and Optimization