Advances in Differential Privacy and Differentially Private Machine Learning

Saswat Das,Subhankar Mishra

DOI: https://doi.org/10.1007/978-981-97-0407-1_7

2024-04-07

Abstract:There has been an explosion of research on differential privacy (DP) and its various applications in recent years, ranging from novel variants and accounting techniques in differential privacy to the thriving field of differentially private machine learning (DPML) to newer implementations in practice, like those by various companies and organisations such as census bureaus. Most recent surveys focus on the applications of differential privacy in particular contexts like data publishing, specific machine learning tasks, analysis of unstructured data, location privacy, etc. This work thus seeks to fill the gap for a survey that primarily discusses recent developments in the theory of differential privacy along with newer DP variants, viz. Renyi DP and Concentrated DP, novel mechanisms and techniques, and the theoretical developments in differentially private machine learning in proper detail. In addition, this survey discusses its applications to privacy-preserving machine learning in practice and a few practical implementations of DP.

Cryptography and Security

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is the lack of a comprehensive review of the latest progress in differential privacy theory, new variants, new technical mechanisms, and the theoretical development of differentially private machine learning (DPML) in the current research literature on differential privacy (DP) and its applications in machine learning. Specifically, although there are already many reviews for specific application scenarios or technical details, these reviews often focus on specific application backgrounds, such as data release, specific machine - learning tasks, unstructured data analysis, location privacy, etc., and there is no review that comprehensively covers the latest progress in differential privacy theory and its applications in machine learning.

To solve this problem, this paper aims to fill this gap and mainly discusses the latest developments in differential privacy theory, including new differential privacy variants (such as Rényi differential privacy and concentrated differential privacy), novel mechanisms and techniques, and the theoretical progress of differentially private machine learning. In addition, this paper also explores the privacy - protecting machine - learning applications of differential privacy in practice and introduces several actual differential privacy implementation cases.

Through such a review, the author hopes to provide readers with a resource for comprehensively understanding the latest progress in differential privacy and its applications in machine learning. Especially for researchers and practitioners who hope to gain in - depth understanding of the technical details of differential privacy, this paper provides important reference value.