RESEARCH AREAS

RESEARCH INTERESTS

• Data Privacy

• Trustworthy Machine Learning

• Federated Learning

RESEARCH PROJECTS

Analytical Framework to Quantify Information Leakage and Memorization in Machine Learning

Machine learning models can "memorize" specific data points from their training data, impacting their predictions and potentially leaking sensitive information. This project aims to understand how this memorization affects models and develop methods to mitigate it.

• Security, Artificial Intelligence

RESEARCH GROUPS

TEACHING INNOVATIONS

SELECTED PUBLICATIONS

• Rishav Chourasia*, Jiayuan Ye*, and Reza Shokri. Differential Privacy Dynamics of Langevin Diffusion and Noisy Gradient Descent. Conference on Neural Information Processing Systems (NeurIPS) - Spotlight, 2021
• Hongyan Chang, and Reza Shokri. On the Privacy Risks of Algorithmic Fairness. IEEE European Symposium on Security and Privacy (EuroSP), 2021
• Reza Shokri, Martin Strobel, and Yair Zick On the Privacy Risks of Model Explanations AAAI/ACM Conference on AI, Ethics, and Society (AIES), 2021
• Sasi Kumar Murakonda, Reza Shokri, and George Theodorakopoulos Quantifying the Privacy Risks of Learning High-Dimensional Graphical Models International Conference on Artificial Intelligence and Statistics (AISTATS), 2021
• Liwei Song, Reza Shokri, and Prateek Mittal. Privacy Risks of Securing Machine Learning Models against Adversarial Examples. ACM Conference on Computer and Communications Security CCS, 2019.
• Milad Nasr, Reza Shokri, and Amir Houmansadr. Comprehensive Privacy Analysis of Deep Learning: Passive and Active White-box Inference Attacks against Centralized and Federated Learning. IEEE Symposium on Security and Privacy S&P -- Oakland, 2019.
• Milad Nasr, Reza Shokri, and Amir HoumansadrMachine Learning with Membership Privacy using Adversarial RegularizationIn the ACM Conference on Computer and Communications Security CCS, 2018
• Reza Shokri, Marco Stronati, Congzheng Song, and Vitaly ShmatikovMembership Inference Attacks against Machine Learning ModelsIn IEEE Symposium on Security and Privacy S&P -- Oakland, 2017.
• Vincent Bindschaedler, Reza Shokri, and Carl GunterPlausible Deniability for Privacy-Preserving Data SynthesisIn the Proceedings of the VLDB Endowment International Conference on Very Large Data Bases PVLDB, 2017.
• Vincent Bindschaedler and Reza Shokri.Synthesizing Plausible Privacy-Preserving Location TracesIn IEEE Symposium on Security and Privacy S&P -- Oakland, 2016.
• Reza Shokri and Vitaly Shmatikov.Privacy-Preserving Deep LearningIn ACM Conference on Computer and Communications Security CCS, 2015.Invited to Annual Allerton Conference on Communication, Control, and Computing Allerton 2015
• Reza Shokri.Privacy Games: Optimal User-Centric Data ObfuscationIn Privacy Enhancing Technologies Symposium PETS, 2015
• Arthur Gervais, Reza Shokri, Adish Singla, Srdjan Capkun, and Vincent Lenders.Quantifying Web-Search Privacy. In ACM Conference on Computer and Communications Security CCS, 2014.
• Reza Shokri, George Theodorakopoulos, Carmela Troncoso, Jean-Pierre Hubaux, and Jean-Yves Le Boudec.Protecting Location Privacy: Optimal Strategy against Localization AttacksIn the 19th ACM Conference on Computer and Communications Security CCS, 2012.
• Reza Shokri, George Theodorakopoulos, Jean-Yves Le Boudec, and Jean-Pierre Hubaux.Quantifying Location PrivacyIn IEEE Symposium on Security and Privacy S&P -- Oakland, 2011.
• Neel Patel, Reza Shokri, and Yair Zick. Model Explanations with Differential Privacy. ACM Conference on Fairness, Accountability, and Transparency (FAccT), 2022
• Hannah Brown, Katherine Lee, Fatemehsadat Mireshghallah, Reza Shokri, and Florian Tramer. What Does it Mean for a Language Model to Preserve Privacy? ACM Conference on Fairness, Accountability, and Transparency (FAccT), 2022
• Fatemehsadat Mireshghallah, Kartik Goyal, Archit Uniyal, Taylor Berg-Kirkpatrick, and Reza Shokri. Quantifying Privacy Risks of Masked Language Models Using Membership Inference Attacks. The Conference on Empirical Methods in Natural Language Processing (EMNLP), 2022
• Florian Tramèr, Reza Shokri, Ayrton San Joaquin, Hoang Le, Matthew Jagielski, Sanghyun Hong, and Nicholas Carlini. Truth Serum: Poisoning Machine Learning Models to Reveal Their Secrets. ACM Conference on Computer and Communications Security (CCS), 2022
• Jiayuan Ye, Aadyaa Maddi, Sasi Kumar Murakonda, Vincent Bindschaedler, and Reza Shokri. Enhanced Membership Inference Attacks against Machine Learning Models. ACM Conference on Computer and Communications Security (CCS), 2022
• Jiayuan Ye and Reza Shokri. Differentially Private Learning Needs Hidden State (Or Much Faster Convergence). Conference on Neural Information Processing Systems (NeurIPS), 2022. Also presented at the Symposium on Foundations of Responsible Computing (FORC), 2022
• Zebang Shen, Jiayuan Ye, Anmin Kang, Hamed Hassani, and Reza Shokri. Share Your Representation Only: Guaranteed Improvement of the Privacy-Utility Tradeoff in Federated Learning. International Conference on Learning Representations (ICLR), 2023
• Hongyan Chang and Reza Shokri. Bias Propagation in Federated Learning. International Conference on Learning Representations (ICLR), 2023

AWARDS & HONOURS

• Asian Young Scientist Fellowship 2023

• NUS School of Computing Faculty Teaching Excellence Award 2023

• IEEE Security and Privacy (S&P) Test-of-Time Award 2021 (Quantifying Location Privacy)

• Facebook (Meta) Faculty Research Award 2021 (Auditing Data Privacy in Machine Learning)

• VMWare Early Career Faculty Award 2021 (Data Privacy and Trustworthy Machine Learning)

• Intel Faculty Research Award 2021 (Privacy-Preserving Federated Learning - Private AI Research Institute)

• Runner-up for the annual PET Award for Outstanding Research in Privacy Enhancing Technologies 2012 (Quantifying Location Privacy)

• NUS Early Career Research Award 2019 (Trustworthy Machine Learning for High-Dimensional Models)

• NUS Presidential Young Professorship 2019 (Privacy in Machine Learning)

• The Caspar Bowden Award for Outstanding Research in Privacy Enhancing Technologies 2018 (Privacy Risks of Machine Learning Models)

COURSES TAUGHT