Research map / Federated learning and privacy / Differential privacy
A Bayesian Nonparametric Framework for Private, Fair, and Balanced Tabular Data Synthesis
Forough Fazeli-Asl, Michael Minyi Zhang, Linglong Kong, Bei Jiang
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 mutual-information fairness regularizer (lambda_F=1) (ablation) | CBNP-VAECGAN · COMPAS | MI(Y,S): 0.0263 -> 1e −5 | Improved |
| add fairness regularizer (ablation) | CBNP-VAECGAN · COMPAS | P(High risk | African-American): 0.146 -> 0.103 | Improved |
| add fairness regularizer (ablation) | CBNP-VAECGAN · COMPAS | P(High risk | Asian): 0.058 -> 0.101 | Mixed |
| add fairness regularizer (ablation) | CBNP-VAECGAN · COMPAS | accuracy (DTC): 0.894 -> 0.886 | Hurt |
| add fairness regularizer (ablation) | CBNP-VAECGAN · COMPAS | F1 (DTC): 0.921 -> 0.903 | Hurt |
| add fairness regularizer (ablation) | CBNP-VAECGAN · COMPAS | MMD: 0.0011 -> 0.0018 | Hurt |
| balance Sex, Marital and Language categories in addition to fairness (ablation) | CBNP-VAECGAN · COMPAS | accuracy (DTC): 0.886 -> 0.883 | No clear effect |
| increase Dirichlet concentration a and number of privatized attributes | CBNP-VAECGAN · Adult | utility (accuracy, F1, MMD) | Hurt |
| use small a such as 10^-6 with strong local privacy | CBNP-VAECGAN · Adult | utility | No clear effect |
| use CBNP-VAECGAN instead of FairGAN, DECAF, TabFairGAN | CBNP-VAECGAN · Adult | fairness and faithfulness | Improved |
| add class balancing on W and MS with lambda_F=1 | CBNP-VAECGAN · Adult | utility and fairness: ACC 79.8%, F1 58.6% | No clear effect |
Closest papers by meaning
- Breaking the Quality–Privacy Tradeoff in Tabular Data Generation via In-Context Learning (NeurIPS 2026 · similarity 0.88)
- CuTS: Customizable Tabular Synthetic Data Generation (ICML 2024 · similarity 0.87)
- Privacy-Preserving Data Release Leveraging Optimal Transport and Particle Gradient Descent (ICML 2024 · similarity 0.87)
- Masked Language Modeling Becomes Conditional Density Estimation for Tabular Data Synthesis (AAAI 2025 · similarity 0.86)
- Do You Really Need Public Data? Surrogate Public Data for Differential Privacy on Tabular Data (NeurIPS 2025 · similarity 0.86)
- HARMONIC: Harnessing LLMs for Tabular Data Synthesis and Privacy Protection (NeurIPS 2024 · 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.
