Research map / Federated learning and privacy / Differential privacy
PAPRIKA: Private Online False Discovery Rate Control
Wanrong Zhang, Gautam Kamath, Rachel Cummings
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 lambda_j = alpha_j (PAPRIKA AI) instead of constant lambda_j=0.2 in the private procedure (ablation) | PAPRIKA · Bernoulli and truncated exponential experiments | power and FDR: e.g. power 0.815 vs 0.785 at pi=0.05, eps=5 (Bernoulli), FDR 0.007 vs 0.053 | Improved |
| Naive Laplace privatization of p-values followed by SAFFRON (LapSAFFRON) instead of PAPRIKA (ablation) | private online FDR baseline · Bernoulli and truncated exponential experiments, eps=5 | power: power about 0.505 vs 0.815 (Bernoulli, pi=0.05) | Hurt |
| Decrease privacy parameter eps from 5 to 3 (ablation) | PAPRIKA AI and PAPRIKA · Bernoulli observations | power: PAPRIKA AI power 0.560 at eps=3 vs 0.815 at eps=5 (pi=0.05) | Hurt |
| Larger shift size s (from 0.5 to 2) in the threshold shift A (ablation) | PAPRIKA · Bernoulli observations, eps=5 | FDR and power | Mixed |
| Use lambda_j=alpha_j with PAPRIKA AI at eps=5 on truncated exponential data, versus SAFFRON AI non-private | PAPRIKA AI · Truncated exponential, pi=0.05 | FDR: FDR 0 vs 0 for SAFFRON AI; private FDR is visually indistinguishable from 0 | No clear effect |
Closest papers by meaning
- SCORE: A Unified Framework for Overshoot Refund in Online FDR Control (ICML 2026 · similarity 0.90)
- Asynchronous Online Testing of Multiple Hypotheses (JMLR 2021 · similarity 0.90)
- ADDIS: an adaptive discarding algorithm for online FDR control with conservative nulls (NeurIPS 2019 · similarity 0.88)
- Generalized Boundary FDR Control under Arbitrary Dependence: An Approach on Closure Principle (ICML 2026 · similarity 0.88)
- Finite Resources False Discovery Rate Control on Structured Hypothesis Spaces (NeurIPS 2026 · similarity 0.87)
- e-GAI: e-value-based Generalized $\alpha$-Investing for Online False Discovery Rate Control (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.
