Privacy aware learning

DOI10.1145/2666468zbMATH Open1321.68400arXiv1210.2085OpenAlexW2151320232MaRDI QIDQ5501941FDOQ5501941

Authors: John C. Duchi, Michael Jordan, Martin J. Wainwright 

Publication date: 14 August 2015

Published in: Journal of the ACM (Search for Journal in Brave)

Abstract: We study statistical risk minimization problems under a privacy model in which the data is kept confidential even from the learner. In this local privacy framework, we establish sharp upper and lower bounds on the convergence rates of statistical estimation procedures. As a consequence, we exhibit a precise tradeoff between the amount of privacy the data preserves and the utility, as measured by convergence rate, of any statistical estimator or learning procedure.

Full work available at URL: https://arxiv.org/abs/1210.2085

Recommendations

• Minimax Optimal Procedures for Locally Private Estimation
• What can we learn privately?
• The cost of privacy: optimal rates of convergence for parameter estimation with differential privacy
• Geometrizing rates of convergence under local differential privacy constraints
• Differentially private empirical risk minimization

zbMATH Keywords

machine learninglower boundsdifferential privacyminimax convergence ratessaddle points

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) Convex programming (90C25) Cryptography (94A60) Computational difficulty of problems (lower bounds, completeness, difficulty of approximation, etc.) (68Q17)

Cites Work

• Asymptotic Statistics
• Title not available (Why is that?)
• Elements of Information Theory
• Acceleration of Stochastic Approximation by Averaging
• Foundations of Modern Probability
• Mirror descent and nonlinear projected subgradient methods for convex optimization.
• Convergence of estimates under dimensionality restrictions
• Robust Stochastic Approximation Approach to Stochastic Programming
• Randomized response: a survey technique for eliminating evasive answer bias
• Differential Privacy: A Survey of Results
• Title not available (Why is that?)
• Title not available (Why is that?)
• Differentially private empirical risk minimization
• Theory of Cryptography
• Uniqueness of solution in linear programming
• Title not available (Why is that?)
• Lectures on Choquet's theorem
• The geometry of differential privacy: the sparse and approximate cases
• On the geometry of differential privacy
• Title not available (Why is that?)
• Efficient noise-tolerant learning from statistical queries
• Information-theoretic determination of minimax rates of convergence
• Title not available (Why is that?)
• Compressed and Privacy-Sensitive Sparse Regression
• What can we learn privately?
• Information-Theoretic Lower Bounds on the Oracle Complexity of Stochastic Convex Optimization
• Estimating Risks of Identification Disclosure in Microdata
• Title not available (Why is that?)
• A statistical framework for differential privacy
• Information theoretic security
• On the Question of Statistical Confidentiality
• Differential privacy and robust statistics
• Differential privacy for functions and functional data
• Privacy-preserving statistical estimation with optimal convergence rates
• Universally utility-maximizing privacy mechanisms

Cited In (21)

• Statistical Query Algorithms for Mean Vector Estimation and Stochastic Convex Optimization
• Optimal locally private estimation under \(\ell_p\) loss for \(1\le p\le 2\)
• Locally differentially private item-based collaborative filtering
• Distributionally-robust machine learning using locally differentially-private data
• Semi-supervised learning with summary statistics
• Learning with differential privacy: stability, learnability and the sufficiency and necessity of ERM principle
• Differentially private empirical risk minimization
• Compressive statistical learning with random feature moments
• Private sequential learning
• On density estimation at a fixed point under local differential privacy
• Strongly universally consistent nonparametric regression and classification with privatised data
• A compressive privacy approach to generalized information bottleneck and privacy funnel problems
• Econometrics with privacy preservation
• Title not available (Why is that?)
• Inferring ground truth from crowdsourced data under local attribute differential privacy
• Geometrizing rates of convergence under local differential privacy constraints
• The complexity of differential privacy
• What can we learn privately?
• Privacy-preserving set-based estimation using partially homomorphic encryption
• Federated Offline Reinforcement Learning
• Perturbation of convex risk minimization and its application in differential private learning algorithms

This page was built for publication: Privacy aware learning

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5501941)