Research map / Federated learning and privacy / Differential privacy
Differentially Private Data Release for Mixed-type Data via Latent Factor Models
Yanqing Zhang, Qi Xu, Niansheng Tang, Annie Qu
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 |
|---|---|---|---|
| Use latent factor model synthetic release (DPFM) instead of PrivBay | differentially private data release · Synthetic setting I (AR-1, eps=0.1) | accuracy: 0.836 -> 0.907 | Improved |
| Use DPFM instead of PrivBay | differentially private data release · Synthetic setting I (AR-1, eps=0.1) | AUC: 0.513 -> 0.921 | Improved |
| Use DPFM instead of PrivBay | differentially private data release · Synthetic setting II (AR-1, eps=0.1) | AUC: 0.503 -> 0.853 | Improved |
| Use DPFM instead of PrivBay | differentially private data release · Census Income (eps=0.1) | accuracy: 0.722 -> 0.785 | Improved |
| Use DPFM instead of PrivBay | differentially private data release · Absenteeism at Work (eps=0.1) | accuracy: 0.511 -> 0.632 | Improved |
| Use DPFM instead of PrivBay | differentially private data release · Breast Cancer (eps=0.1) | accuracy: 0.527 -> 0.678 | Improved |
| Allocate privacy budget as (2eps/3, eps/6, eps/6) instead of equal thirds (ablation) | DPFM · Synthetic setting I (AR-1, eps=0.1) | accuracy: 0.907 -> 0.904 | No clear effect |
| Allocate privacy budget as (eps/6, eps/6, 2eps/3) instead of equal thirds (ablation) | DPFM · Synthetic setting I (AR-1, eps=0.1) | accuracy: 0.907 -> 0.906 | No clear effect |
| Increase privacy budget from eps=0.1 to eps=5 (ablation) | DPFM1 · Synthetic setting I (AR-1) | accuracy: 0.907 -> 0.906 | No clear effect |
| Increase privacy budget from eps=0.1 to eps=5 (ablation) | DPFM · Census Income | accuracy: 0.785 -> 0.785 | No clear effect |
| Increase privacy budget from eps=0.1 to eps=5 (ablation) | PrivBay · Synthetic setting II (AR-1) | AUC: 0.503 -> 0.584 | Mixed |
Closest papers by meaning
- A Privacy-Preserving Framework for Generative Model-driven Synthetic Datasets (AAAI 2025 · similarity 0.87)
- Privacy Amplification Through Synthetic Data: Insights from Linear Regression (ICML 2025 · similarity 0.87)
- Privacy Amplification Persists under Unlimited Synthetic Data Release (NeurIPS 2026 · similarity 0.87)
- Contrastive Private Data Synthesis via Weighted Multi-PLM Fusion (ICML 2025 · similarity 0.86)
- Differentially Private Sum-Product Networks (ICML 2024 · similarity 0.86)
- Does Training with Synthetic Data Truly Protect Privacy? (ICLR 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.
