Research map / Federated learning and privacy / Differential privacy
Privately Learning Markov Random Fields
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Apply approximate DP rather than no privacy to structure learning | Ising model structure learning · Ising model | sample complexity dependence on p: Θ(log{p}) -> Θ(log{p}) | No clear effect |
| Apply approximate DP to parameter learning | Ising model parameter learning · Ising model | sample complexity dependence on p: Θ(log{p}) -> Θ(√p) | Hurt |
| Tighten privacy from approximate DP to zero-concentrated DP for structure learning | Ising model structure learning · Ising model | sample complexity dependence on p: Θ(log{p}) -> Θ(√p) | Hurt |
| Tighten privacy from zero-concentrated DP to pure DP for structure learning | Ising model structure learning · Ising model | sample complexity dependence on p: Θ(√p) -> Ω(p) | Hurt |
| Use private parameter learning instead of structure learning | Approximate-DP Ising model learning · Ising model | sample complexity dependence on p: Θ(log{p}) -> Θ(√p) | Hurt |
| Replace PMW with sepFEM for parity-query release | Binary t-wise MRF maximal-monomial learning · Binary t-wise MRF | sample complexity dependence on p in the parity-query term: √p -> p^{5/4} | Hurt |
| Use sepFEM instead of PMW for parity-query release | Binary t-wise MRF maximal-monomial learning · Binary t-wise MRF | runtime: exponential in p -> polynomial time | Helped |
Closest papers by meaning
- Certification for Differentially Private Prediction in Gradient-Based Training (ICML 2025 · similarity 0.89)
- Tractable MCMC for Private Learning with Pure and Gaussian Differential Privacy (ICLR 2024 · similarity 0.88)
- Exactly Minimax-Optimal Locally Differentially Private Sampling (NeurIPS 2024 · similarity 0.88)
- The Limits of Differential Privacy in Online Learning (NeurIPS 2024 · similarity 0.88)
- Public-data Assisted Private Stochastic Optimization: Power and Limitations (NeurIPS 2024 · similarity 0.88)
- Oracle-Efficient Differentially Private Learning with Public Data (NeurIPS 2024 · similarity 0.88)
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.