Abstract:Local Differential Privacy (LDP) can provide each user with strong privacy guarantees under untrusted data curators while ensuring accurate statistics derived from privatized data. Due to its powerfulness, LDP has been widely adopted to protect privacy in various tasks (e.g., heavy hitters discovery, probability estimation) and systems (e.g., Google Chrome, Apple iOS). In particular, <math>(∊,δ)</math>-LDP has been studied in related statistical tasks like private learning and hypothesis testing, but is mainly achieved by using Gaussian mechanism, leading to the limited data utility. In this paper, we investigate several novel mechanisms that achieve <math>(∊,δ)</math>-LDP with higher data utility in collecting and analyzing users' data. Specifically, we first design two <math>(∊,δ)</math>-LDP algorithms for mean estimations on multi-dimensional numeric data, which can ensure higher accuracy than the optimal Gaussian mechanism. Then, we investigate different local protocols for frequency estimations on categorical attributes under <math>(∊,δ)</math>-LDP. Based on the proposed mechanisms, we further study on <math>(∊,δ)</math>-LDP-compliant stochastic gradient descent algorithms for machine learning models. Besides, the theoretical analysis of the error bound and the variance of the proposed algorithms are also presented in the paper. We have conducted extensive experiments on both real-world and synthetic datasets and demonstrated the high data utility of our proposed algorithms in the perspectives of simple data statistics tasks and complex machine learning tasks. The experimental results have shown that our proposed algorithms can effectively improve the data utility in different tasks while alleviating the privacy concerns of each individual.

The Trade-off Between Privacy and Utility in Local Differential Privacy

Utility Analysis and Enhancement of LDP Mechanisms in High-Dimensional Space

A Quantifying Method for Trade-off Between Privacy and Utility

Local Differential Privacy for data collection and analysis

Not one but many Tradeoffs: Privacy Vs. Utility in Differentially Private Machine Learning

A Privacy-Preserving Game Model for Local Differential Privacy by Using Information-Theoretic Approach

A Utility-Optimized Framework for Personalized Private Histogram Estimation

Utility–Privacy Trade-Off in Distributed Machine Learning Systems

Key-Value Data Collection with Distribution Estimation under Local Differential Privacy

Improving Data Utility Through Game Theory in Personalized Differential Privacy.

LDPTube: Theoretical Utility Benchmark and Enhancement for LDP Mechanisms in High-dimensional Space

On robustness and local differential privacy

Learning with User-Level Local Differential Privacy

On the Tradeoff Between Privacy and Utility in Data Publishing.

On the Tradeoff Between Data-Privacy and Utility for Data Publishing

A Distributed Privacy-Utility Tradeoff Method Using Distributed Lossy Source Coding with Side Information

Minimax Optimal Two-Sample Testing under Local Differential Privacy

Utility-Preserving Privacy Mechanisms for Counting Queries

Privacy-Preserving Monotonicity of Differential Privacy Mechanisms

Meeting Utility Constraints in Differential Privacy: A Privacy-Boosting Approach

Local privacy and statistical minimax rates