Differentially Private Wasserstein Barycenters
Not on the map. Why?
Abstract (arXiv)
The Wasserstein barycenter is defined as the mean of a set of probability measures under the optimal transport metric, and has numerous applications spanning machine learning, statistics, and computer graphics. In practice these input measures are empirical distributions built from sensitive datasets, motivating a differentially private (DP) treatment. We present, to our knowledge, the first algorithms for computing Wasserstein barycenters under differential privacy. Empirically, on synthetic data, MNIST, and large-scale U.S. population datasets, our methods produce high-quality private barycenters with strong accuracy-privacy tradeoffs.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 31 of its 72 references that are in the map. First posted on arXiv 3 Oct 2025.
Builds on
31 of 72 references.
References · 31
- 2013Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distancescited by 2,316
- 2018Privacy Amplification by Subsampling: Tight Analyses via Couplings and Divergencescited by 242
- 2018Differentially Private Generative Adversarial Networkcited by 235
- 2017Near-linear time approximation algorithms for optimal transport via Sinkhorn iterationcited by 223
- 2013Fast Computation of Wasserstein Barycenterscited by 216
- 2018Amplification by Shuffling: From Local to Central Differential Privacy via Anonymitycited by 164
- 2017Collecting Telemetry Data Privatelycited by 149
- 2018Distributed Differential Privacy via Shufflingcited by 115
- 2021Winning the NIST Contest: A scalable and general approach to differentially private synthetic datacited by 102
- 2019Statistical bounds for entropic optimal transport: sample complexity and the central limit theoremcited by 85
- 2018Performance of Johnson-Lindenstrauss Transform for k-Means and k-Medians Clusteringcited by 53
- 2020Differentially Private Clustering: Tight Approximation Ratioscited by 34
19 more
- 2020LinkedIn's Audience Engagements API: A Privacy Preserving Data Analytics System at Scalecited by 26
- 2018Stochastic Wasserstein Barycenterscited by 25
- 2020Wasserstein barycenters can be computed in polynomial time in fixed dimensioncited by 23
- 2018Differentially Private k-Means with Constant Multiplicative Errorcited by 22
- 2024Differentially Private Release of Israel's National Registry of Live Birthscited by 17
- 2021Smooth p-Wasserstein Distance: Structure, Empirical Approximation, and Statistical Applicationscited by 15
- 2022Improved Approximations for Euclidean k-means and k-median, via Nested Quasi-Independent Setscited by 14
- 2021Dimensionality Reduction for Wasserstein Barycentercited by 13
- 2020k-Variance: A Clustered Notion of Variancecited by 11
- 2023Algorithmically Effective Differentially Private Synthetic Datacited by 11
- 2020Differentially Private Clustering via Maximum Coveragecited by 9
- 2024Instance-Optimal Private Density Estimation in the Wasserstein Distancecited by 9
- 2020Differentially private k-means clustering via exponential mechanism and max covercited by 8
- 2021Convergence of Gaussian-smoothed optimal transport distance with sub-gamma distributions and dependent samplescited by 8
- 2021An Improved Local Search Algorithm for k-Mediancited by 8
- 2022Private Query Release via the Johnson-Lindenstrauss Transformcited by 8
- 2020Multimarginal Wasserstein Barycenter for Stain Normalization and Augmentationcited by 6
- 2021Randomized Dimensionality Reduction for Facility Location and Single-Linkage Clusteringcited by 6
- 2022The Johnson-Lindenstrauss Lemma for Clustering and Subspace Approximation: From Coresets to Dimension Reductioncited by 5
Built on by
Citing papers per year: 2026: 2.
2 papers.