← Authors

Abhradeep Guha Thakurta

28 papers · 2013–2025 · 5 conferences · across top CS/AI conferences

Achievements

🐝 Cross-Pollinator (14) 🌈 Renaissance Researcher (5) 🐣 Hot Topic Early Bird πŸŒ‰ Interdisciplinary Bridge 🌍 Conference Polyglot (5) πŸƒ Academic Marathon (12) πŸ—ΊοΈ Taxonomy Completionist (30) 🧭 Keyword Pioneer πŸ”¬ Deep Specialist (19) πŸ‘‘ Triple Crown πŸ† Keyword Champion (2) 🌱 Topic Pioneer πŸ—ƒοΈ Keyword Collector (82) ⚑ Prolific Year (8) πŸ“ˆ Trend Setter πŸ’Ž Century Club (28) πŸ”₯ Unstoppable (9) ❓ The Questioner (2)

Conferences

NIPS (15) ICML (6) ICLR (5) ALT (1) COLT (1)

Top co-authors

Arun Ganesh (9) Adam Smith (7) Christopher A. Choquette-Choo (7) Thomas Steinke (6) Shuang Song (4) Om Thakkar (4) Daogao Liu (3) Raef Bassily (3) Prateek Jain (3) John Rush (3)

Research topics

Differential Privacy (6)

Keywords

differential privacy (19) stochastic gradient descent (4) federated learning (3) excess risk (2) private training (2) online learning (2) convex optimization (2) user-level privacy (2) regret bound (2) matrix factorization (2) lasso estimator (2) feature selection (2) privacy amplification (2) sample complexity (1) uncertainty quantification (1) logistic regression (1) model selection (1) algorithmic stability (1) l1 regularization (1) vc dimension (1)

Papers

Near-Exact Privacy Amplification for Matrix Mechanisms ICLR 2025
Near-Optimal Rates for O(1)-Smooth DP-SCO with a Single Epoch and Large Batches ALT 2025
The Last Iterate Advantage: Empirical Auditing and Principled Heuristic Analysis of Differentially Private SGD ICLR 2025
Privacy Amplification for Matrix Mechanisms ICLR 2024
Correlated Noise Provably Beats Independent Noise for Differentially Private Learning ICLR 2024
Private Gradient Descent for Linear Regression: Tighter Error Bounds and Instance-Specific Uncertainty Estimation ICML 2024
Improved Differentially Private and Lazy Online Convex Optimization: Lower Regret without Smoothness Requirements ICML 2024
Faster Differentially Private Convex Optimization via Second-Order Methods NIPS 2023
Training Private Models That Know What They Don’t Know NIPS 2023
Private (Stochastic) Non-Convex Optimization Revisited: Second-Order Stationary Points and Excess Risks NIPS 2023
(Amplified) Banded Matrix Factorization: A unified approach to private training NIPS 2023
Measuring Forgetting of Memorized Training Examples ICLR 2023
Multi-Epoch Matrix Factorization Mechanisms for Private Machine Learning ICML 2023
Why Is Public Pretraining Necessary for Private Model Training? ICML 2023
Multi-Task Differential Privacy Under Distribution Skew ICML 2023
Improved Differential Privacy for SGD via Optimal Private Linear Operators on Adaptive Streams NIPS 2022
When Does Differentially Private Learning Not Suffer in High Dimensions? NIPS 2022
A Separation Result Between Data-oblivious and Data-aware Poisoning Attacks NIPS 2021
Differentially Private Model Personalization NIPS 2021
Privacy Amplification via Random Check-Ins NIPS 2020
The Flajolet-Martin Sketch Itself Preserves Differential Privacy: Private Counting with Minimal Space NIPS 2020
Private Stochastic Convex Optimization with Optimal Rates NIPS 2019
Model-Agnostic Private Learning NIPS 2018
Practical Locally Private Heavy Hitters NIPS 2017
Nearly Optimal Private LASSO NIPS 2015
(Near) Dimension Independent Risk Bounds for Differentially Private Learning ICML 2014
Differentially Private Feature Selection via Stability Arguments, and the Robustness of the Lasso COLT 2013
(Nearly) Optimal Algorithms for Private Online Learning in Full-information and Bandit Settings NIPS 2013