Recent Advances of Differential Privacy in Centralized Deep Learning: A Systematic Survey

TL;DR

This survey reviews recent progress in applying differential privacy to centralized deep learning, highlighting advancements, challenges, and future directions for ensuring data privacy in machine learning models.

Contribution

It systematically analyzes recent developments, open problems, and potential future research areas in differentially private centralized deep learning.

Findings

Advances in privacy-utility trade-offs

Methods for auditing and evaluating private models

Protection against various threats and attacks

Abstract

Differential Privacy has become a widely popular method for data protection in machine learning, especially since it allows formulating strict mathematical privacy guarantees. This survey provides an overview of the state-of-the-art of differentially private centralized deep learning, thorough analyses of recent advances and open problems, as well as a discussion of potential future developments in the field. Based on a systematic literature review, the following topics are addressed: auditing and evaluation methods for private models, improvements of privacy-utility trade-offs, protection against a broad range of threats and attacks, differentially private generative models, and emerging application domains.