mlsanityData to 7 Oct 2026

Online Differentially Private Synthetic Data Generation

arXiv 2402.08012 ↗First posted 12 Feb 2024math.ST cs.DS cs.LGcited by 4

Not on the map. Why?

Abstract (arXiv)

We present a polynomial-time algorithm for online differentially private synthetic data generation. For a data stream within the hypercube [0,1]d and an infinite time horizon, we develop an online algorithm that generates a differentially private synthetic dataset at each time t. This algorithm achieves a near-optimal accuracy bound of O(log(t)t-1/d) for d≥ 2 and O(log4.5(t)t-1) for d=1 in the 1-Wasserstein distance. This result extends the previous work on the continual release model for counting queries to Lipschitz queries. Compared to the offline case, where the entire dataset is available at once, our approach requires only an extra polylog factor in the accuracy bound.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 28 of its 66 references that are in the map. First posted on arXiv 12 Feb 2024.

Builds on

28 of 66 references.

References · 28

  1. 2021Practical and Private (Deep) Learning without Sampling or Shufflingcited by 169
  2. 2021A Universal Law of Robustness via Isoperimetrycited by 150
  3. 2016Minimax Optimal Procedures for Locally Private Estimationcited by 103
  4. 2021A Dynamical System Perspective for Lipschitz Neural Networkscited by 59
  5. 2022Private Synthetic Data for Multitask Learning and Marginal Queriescited by 35
  6. 2021The Price of Differential Privacy under Continual Observationcited by 32
  7. 2018Optimizing error of high-dimensional statistical queries under differential privacycited by 29
  8. 2015Differentially Private k-Means Clusteringcited by 27
  9. 2016Make Up Your Mind: The Price of Online Queries in Differential Privacycited by 26
  10. 2022Almost Tight Error Bounds on Differentially Private Continual Countingcited by 25
  11. 2022Constant matters: Fine-grained Complexity of Differentially Private Continual Observationcited by 23
  12. 2013Efficient Algorithms for Privately Releasing Marginals via Convex Relaxationscited by 17
16 more
  1. 2023Differentially Private Neural Tangent Kernels for Privacy-Preserving Data Generationcited by 15
  2. 2023A Unifying Framework for Differentially Private Sums under Continual Observationcited by 15
  3. 2021Differentially Private Histograms under Continual Observation: Streaming Selection into the Unknowncited by 14
  4. 2023Algorithmically Effective Differentially Private Synthetic Datacited by 11
  5. 2023Differentially Private Continual Releases of Streaming Frequency Moment Estimationscited by 10
  6. 2018Differential Privacy for Growing Databasescited by 8
  7. 2021Private sampling: a noiseless approach for generating differentially private synthetic datacited by 8
  8. 2024Privacy-preserving data release leveraging optimal transport and particle gradient descentcited by 8
  9. 2021Covariance's Loss is Privacy's Gain: Computationally Efficient, Private and Accurate Synthetic Datacited by 7
  10. 2023Certified private data release for sparse Lipschitz functionscited by 7
  11. 2023Continual Release of Differentially Private Synthetic Data from Longitudinal Data Collectionscited by 7
  12. 2023Optimal Bounds on Private Graph Approximationcited by 6
  13. 2022Continual Mean Estimation Under User-Level Privacycited by 5
  14. 2018Differentially Private Continual Release of Graph Statisticscited by 4
  15. 2023Differentially Private Low-dimensional Synthetic Data from High-dimensional Datasetscited by 4
  16. 2024An Algorithm for Streaming Differentially Private Datacited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 3; 2026: 1.

4 papers.

  1. 2025InvisibleInk: High-Utility and Low-Cost Text Generation with Differential Privacycited by 10
  2. 2025Private Evolution Convergescited by 2
  3. 2025Private Continuous-Time Synthetic Trajectory Generation via Mean-Field Langevin Dynamicscited by 0
  4. 2026Minimax optimal differentially private synthetic data for smooth queriescited by 0