Thomas Steinke

Interactive fingerprinting codes and the hardness of preventing false discovery

2016

We show an essentially tight bound on the number of adaptively chosen statistical queries that a computationally efficient algorithm can answer accurately given n samples from an unknown distribution. A statistical query asks for the …

Average-Case Averages: Private Algorithms for Smooth Sensitivity and\n Mean Estimation

2019 · arXiv (Cornell University)

The simplest and most widely applied method for guaranteeing differential\nprivacy is to add instance-independent noise to a statistic of interest that is\nscaled to its global sensitivity. However, global sensitivity is a worst-case\nnotion that is often …

New Oracle-Efficient Algorithms for Private Synthetic Data Release

2020 · arXiv (Cornell University)

We present three new algorithms for constructing differentially private synthetic data---a sanitized version of a sensitive dataset that approximately preserves the answers to a large collection of statistical queries. All three algorithms are \emph{oracle-efficient} in …

Evading Curse of Dimensionality in Unconstrained Private GLMs via Private Gradient Descent

2020 · arXiv (Cornell University)

We revisit the well-studied problem of differentially private empirical risk minimization (ERM). We show that for unconstrained convex generalized linear models (GLMs), one can obtain an excess empirical risk of $\tilde O\left(\sqrt{\texttt{rank}}/εn\right)$, where ${\texttt{rank}}$ is …

Privacy Amplification for Matrix Mechanisms

2023 · arXiv (Cornell University)

Privacy amplification exploits randomness in data selection to provide tighter differential privacy (DP) guarantees. This analysis is key to DP-SGD's success in machine learning, but, is not readily applicable to the newer state-of-the-art algorithms. This …