Research map / Federated learning and privacy / Differential privacy
Post-processing Private Synthetic Data for Improving Utility on Selected Measures
Hao Wang, Shivchander Sudalairaj, John Henning, Kristjan Greenewald, Akash Srivastava
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add post-processing resampling | DPCTGAN · Adult eps=2 | F1 score: 0.22 -> 0.33 | Improved |
| add post-processing resampling | PATECTGAN · Adult eps=4 | F1 score: 0.16 -> 0.49 | Improved |
| add post-processing resampling | PATECTGAN · Mushroom eps=4 | F1 score: 0.29 -> 0.78 | Improved |
| add post-processing resampling | PATECTGAN · Mushroom eps=2 | F1 score: 0.6 -> 0.86 | Improved |
| add post-processing resampling | AIM · Adult eps=2 | F1 score: 0.61 -> 0.61 | No clear effect |
| add post-processing resampling | MST · Mushroom eps=2 | F1 score: 0.83 -> 0.83 | No clear effect |
| add post-processing resampling | GEM · home-credit eps=2 | F1 score: 0.19 -> 0.19 | No clear effect |
| add post-processing resampling | DPCTGAN · Adult eps=2 | Jensen-Shannon distance: 0.07 -> 0.03 | Improved |
| add post-processing resampling | PATECTGAN · Adult eps=2 | Inverse KL: 0.72 -> 0.87 | Improved |
| add post-processing resampling | GEM · home-credit eps=2 | Inverse KL: 0.93 -> 0.97 | Improved |
| add post-processing resampling | GEM · home-credit eps=4 | Inverse KL: 0.95 -> 0.99 | Improved |
| add post-processing resampling | GEM · home-credit | JS distance: 0.01 -> 0.01 | No clear effect |
Closest papers by meaning
- Private Evolution Converges (NeurIPS 2025 · similarity 0.88)
- Debiasing Synthetic Data Generated by Deep Generative Models (NeurIPS 2024 · similarity 0.87)
- Does Training with Synthetic Data Truly Protect Privacy? (ICLR 2025 · similarity 0.87)
- PCEvolve: Private Contrastive Evolution for Synthetic Dataset Generation via Few-Shot Private Data and Generative APIs (ICML 2025 · similarity 0.87)
- Privacy Amplification Through Synthetic Data: Insights from Linear Regression (ICML 2025 · similarity 0.87)
- Analysis of Differentially Private Synthetic Data: A Measurement Error Approach (AAAI 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
