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.

NetDP: In-Network Differential Privacy for Large-Scale Data Processing

Data-Driven Small Cell Placement Optimization With Users' Differential Privacy For Wireless Ngns

Data-Driven Caching with Users' Local Differential Privacy in Information-Centric Networks

Industrial Scale Privacy Preserving Deep Neural Network

A Distributed Privacy-Preserving Framework for Deep Learning with Edge-Cloud Computing.

Differential Privacy Enabled Deep Neural Networks for Wireless Resource Management

A Decentralized Mechanism Based on Differential Privacy for Privacy-Preserving Computation in Smart Grid

When Deep Learning Meets Differential Privacy: Privacy,Security, and More

PrivData Network: A Privacy-preserving On-Chain Data Factory and Trading Market

Over-the-Air Collaborative Inference with Feature Differential Privacy

Local Differential Privacy for Deep Learning

DPI: Ensuring Strict Differential Privacy for Infinite Data Streaming

Computationally efficient data synthesis for AC-OPF: Integrating Physics-Informed Neural Network solvers and active learning

Differentially Private Fine-Tuning of Diffusion Models

The Normal Distributions Indistinguishability Spectrum and its Application to Privacy-Preserving Machine Learning

Local Differential Privacy for data collection and analysis

AdaPDP: Adaptive Personalized Differential Privacy.

DPShield: Optimizing Differential Privacy for High-Utility Data Analysis in Sensitive Domains

E-DPNCT: An Enhanced Attack Resilient Differential Privacy Model For Smart Grids Using Split Noise Cancellation

PP-Stream: Toward High-Performance Privacy-Preserving Neural Network Inference via Distributed Stream Processing.

Scalable Differential Privacy Mechanisms for Real-Time Machine Learning Applications