Random Differential Privacy

By Rob Hall, Alessandro Rinaldo, Larry A. Wasserman. 2011

96 citations recorded by the source corpus.

We propose a relaxed privacy definition called {\em random differential privacy} (RDP). Differential privacy requires that adding any new observation to a database will have small effect on the output of the data-release procedure. Random differential privacy requires that adding a {\em randomly drawn new observation} to a database will have small effect on the output. We show an analog of the composition property of differentially private procedures which applies to our new definition. We show how to release an RDP histogram and we show that RDP histograms are much more accurate than histograms obtained using ordinary differential privacy. We finally show an analog of the global sensitivity framework for the release of functions under our privacy definition.

Subfields: Artificial Intelligence, Electrical and Electronic Engineering

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 Electrical and Electronic Engineering

Citations in the corpus

Privacy Aware Learning (2012, background): “[15] gives sharp minimax rates of convergence for differentially private histogram estimation.”
Minimax Optimal Procedures for Locally Private Estimation (2016, background): “More recent work aims to link differential privacy to statistical concerns (Dwork and Lei 2009; Wasserman and Zhou 2010; Hall, Rinaldo, and Wasserman 2012; Smith 2011; Chaudhuri, Monteleoni, and Sarwate 2011; Rubinstein et al. 2012); in particular, researchers have developed algorithms for private…”
Local Privacy, Data Processing Inequalities, and Statistical Minimax Rates (2013, background): “Similar results hold for histogram estimation [33], classification problems [15], and classical point estimation problems [48]; in this sense, local and global forms of differential privacy can be rather different.”
Finite Sample Differentially Private Confidence Intervals (2017, background): “…lower bounds” for calculating empirical statistics under differential privacy (Hardt and Talwar, 2010; Beimel et al., 2010; Wasserman and Zhou, 2010; Hall et al., 2011), and is also a generalization of the “secrecy of the sample” property of differential privacy (Kasiviswanathan et al., 2011;…”
Pufferfish (2014, background): “P uffer F ish( S , pairs ) satisﬁes the axioms of convexity and transformation invariance.”
Pain-Free Random Differential Privacy with Sensitivity Sampling (2017, methodology): “The following deﬁnition due to Hall et al. (2012) relaxes the requirement that uniform smoothness of response distribution holds on all pairs of databases, to the requirement that uniform smoothness holds for likely database pairs.”
Bayesian Differential Privacy through Posterior Sampling (2013, background): “, the posterior ξ is (2L, 0, exp(−c(L− L0)))-randomly differentially private (Hall et al., 2011) under pseudo-metric ρ.”
A ug 2 01 4 Local Privacy , Data Processing Inequalities , and Minimax Rates (2018, background): “Several researchers have considered quantities similar to our minimax criteria under local (2) or non-local (46) differential privacy [7, 35, 33, 18].”
Coupled-Worlds Privacy: Exploiting Adversarial Uncertainty in Statistical Data Privacy (2013, background): “The goal of this paper is to develop rigorous definitions of privacy for statistical databases that allow us to reason about and exploit existing adversarial uncertainty about the underlying data.”
Generalized Gaussian Mechanism for Differential Privacy (2016, background): “(1), there are softer versions of DP, including the ð ; dÞ-approximate DP [13], the ð ; dÞ-probabilistic DP [14], the ð ; dÞ-random DP (rDP) [15], and the ð ; tÞ-concentrated DP (cDP) [16].”