Research map / Federated learning and privacy / Differential privacy
Relaxed Marginal Consistency for Differentially Private Query Answering
Ryan McKenna, Siddhant Pradhan, Daniel Sheldon, Gerome Miklau
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 |
|---|---|---|---|
| replace exact inference with a saturated region graph | Prox-PGM · Synthetic 100-dimensional data, random and greedy 3-way marginals | measured cliques handled: about 50; 1000 -> 10000 | Improved |
| use region-graph reconstruction | HDMM · fire, 455 random 3-way marginals | RMSE: 3 \times reduction | Improved |
| use a saturated region graph instead of exact inference (ablation) | Prox-PGM · Synthetic 8-dimensional data, in-model 3-way marginals | L1 error | Hurt |
| use a factor graph instead of exact inference (ablation) | Prox-PGM · Synthetic 8-dimensional data, in-model 3-way marginals | L1 error | Hurt |
| use a factor graph instead of a saturated region graph (ablation) | Prox-PGM · Synthetic 8-dimensional data, out-of-model marginals | L1 error | Hurt |
| use a saturated region graph instead of exact inference (ablation) | Prox-PGM · Synthetic 8-dimensional data, out-of-model marginals | L1 error | Hurt |
| use APPGM rather than Private-PGM | MWEM · fire, all 2-way marginals | error at a fixed number of rounds | Hurt |
| run MWEM with APPGM to completion rather than stop at Private-PGM's minimum-error point | MWEM · Additional MWEM experiments | error: 3× as many rounds; 3× as much privacy budget | Improved |
| use APPGM to reconstruct FEM answers | FEM · Adult | error | Improved |
| use an over-saturated region graph to include out-of-model marginals (ablation) | APPGM · Out-of-model marginal estimation | estimate quality | Improved |
| use an over-saturated region graph to include out-of-model marginals (ablation) | APPGM · Out-of-model marginal estimation | runtime | Hurt |
| warm start Convex-GBP messages and run one inner iteration per Prox-PGM iteration (ablation) | APPGM · Practical optimization | objective progress | Improved |
Closest papers by meaning
- Efficient and Private Marginal Reconstruction with Local Non-Negativity (NeurIPS 2024 · similarity 0.91)
- Perturb-and-Project: Differentially Private Similarities and Marginals (ICML 2024 · similarity 0.89)
- Optimality of Matrix Mechanism on $\ell_p^p$-metric (ICLR 2025 · similarity 0.88)
- Computationally Efficient Replicable Learning of Parities and Applications (NeurIPS 2026 · similarity 0.87)
- Private Prediction via Shrinkage (NeurIPS 2026 · similarity 0.87)
- Certification for Differentially Private Prediction in Gradient-Based Training (ICML 2025 · similarity 0.86)
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.
