4942_zakynthinou-2025-square-300x300.jpg

Lydia Zakynthinou

Differential Privacy Privacy-Preserving Machine Learning Robustness In Learning Algorithms Generalization In Statistical Learning Memorization In Machine Learning
Johns Hopkins University Department of Computer Science

About

Lydia Zakynthinou is an assistant professor in the Department of Computer Science and a member of the Johns Hopkins Data Science and AI Institute. She works on the theoretical foundations of trustworthy and reliable machine learning and statistics. Her research focuses on developing privacy-preserving methods—particularly those satisfying the formal definition of differential privacy—and understanding their fundamental limitations. Zakynthinou’s work also explores other dimensions of reliability with surprising connections to privacy, including robustness, generalization, and memorization in learning algorithms. Her goal is to provide safe, principled alternatives that offer formal guarantees as well as competitive performance, empowering data practitioners to adopt trustworthy methods in practice. Zakynthinou has published research in leading machine learning theory venues such as the Conference on Learning Theory, the Conference on Neural Information Processing Systems, and the International Conference on Machine Learning, with select papers recognized by spotlight presentations. Her graduate and post-graduate work has been supported by a Meta Research PhD Fellowship, a Northeastern University Khoury College of Computer Sciences PhD Research Award, and a Foundations of Data Science Institute postdoctoral fellowship. Zakynthinou additionally serves on the program committees of multiple conferences in machine learning, security, and privacy, and is a workshop committee member of the Learning Theory Alliance. She completed her PhD in computer science at Northeastern University, where she was advised by Jonathan Ullman and Huy Lê Nguyễn; her master’s in logic, algorithms, and theory of computation at the National Kapodistrian University of Athens; and her diploma in electrical and computer engineering at the National Technical University of Athens. Before joining Johns Hopkins, Zakynthinou was a postdoctoral research fellow at the Simons Institute for the Theory of Computing at the University of California, Berkeley, hosted by Michael I. Jordan.

Research Performance Summary

15
Total Papers
600
Total Citations
9
H-Index
2018-2025
Active Research Span

First Recorded Paper

Improved algorithms for collaborative PAC learning

Year: 2018

Citations: 44

Venue: Advances in Neural Information Processing Systems 31

Latest Recorded Paper

Tukey Depth Mechanisms for Practical Private Mean Estimation

Year: 2025

Citations: 1

Venue: arXiv preprint arXiv:2502.18698

Last 10 Years Publication Activity

This timeline shows the professor's yearly publication activity.

2025
2
2024
1
2023
3
2021
4
2020
4
2018
1

Publication Venues and Collaboration

Journal, Conference, and Book Publication Breakdown

Conference 8
Journal 0
Book 1

Top Coauthors

Zakynthinou Ullman Steinke

Research Impact by Period

2025-2026

Period Stats
Papers 2
Citations 4
Avg. Citations / Paper 2
H-Index 1

2020-2024

Period Stats
Papers 12
Citations 552
Avg. Citations / Paper 46
H-Index 8

2015-2019

Period Stats
Papers 1
Citations 44
Avg. Citations / Paper 44
H-Index 1

Contact and Professional Links

Contact Information

lzakynthinou@jhu.edu

Detected Research Keywords

Private Mean Estimation Conditional Mutual Information