Research map / Probabilistic and scientific modeling / Molecules and proteins
False Discovery Proportion control for aggregated Knockoffs
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 |
|---|---|---|---|
| Use KOPI with calibrated FDP bounds and harmonic-mean aggregation instead of Vanilla Knockoffs | Knockoff variable selection · Simulated data across varied settings | FDP control | Helped |
| Use KOPI rather than existing FDR-controlling aggregation methods | Knockoff variable selection · Simulated data | power | Helped |
| Use KOPI rather than Closed Testing | Knockoff variable selection · Simulated data | power | Helped |
| Use KOPI rather than Vanilla Knockoffs | Conditional region selection · HCP motor foot | selected regions: 17 regions -> 24 regions | Helped |
| Use KOPI rather than Vanilla Knockoffs | Conditional region selection · HCP emotional | selected regions: 22 regions -> 37 regions | Helped |
| Use KOPI rather than Vanilla Knockoffs | Conditional region selection · HCP relational | selected regions: 58 regions -> 24 regions | Mixed |
| Use KOPI rather than Vanilla Knockoffs | Knockoff gene selection · Lymphomatic leukemia mutation | genes selected in >90% of runs: 0 -> 4 | Helped |
| Use KOPI rather than Vanilla Knockoffs | Knockoff gene selection · Lymphomatic leukemia mutation | spurious detections (<50% of runs): 18 -> 2 | Helped |
| Use harmonic mean instead of arithmetic mean aggregation (ablation) | KOPI · Simulated data, rho = 0.5, 0.6, 0.7 | power: 0.77 -> 0.91; 0.58 -> 0.83; 0.39 -> 0.72 | Helped |
| Use harmonic mean instead of geometric mean aggregation (ablation) | KOPI · Simulated data, rho = 0.5, 0.6, 0.7 | power: 0.87 -> 0.91; 0.77 -> 0.83; 0.61 -> 0.72 | Helped |
| Use harmonic mean instead of quantile aggregation (ablation) | KOPI · Simulated data, rho = 0.5, 0.6, 0.7 | power: 0.90 vs 0.91; 0.83 vs 0.83; 0.72 vs 0.72 | No clear effect |
| Replace harmonic mean aggregation with arithmetic, geometric, or quantile aggregation (ablation) | KOPI · Simulated data, rho = 0.5, 0.6, 0.7 | FDP bound non-coverage: Harmonic: 10 %, 2 %, 2 %; Arithmetic: 0 %, 0 %, 0 %; Geometric: 2 %, 0 %, 0 %; Quantile: 10 %, 4 %, 0 % | No clear effect |
Closest papers by meaning
- Integral-based Knockoffs Inference for Partially Linear Models (AAAI 2026 · similarity 0.89)
- Conformalized Multiple Testing after Data-dependent Selection (NeurIPS 2024 · similarity 0.84)
- Knockoffs Inference for Partially Linear Models with Automatic Structure Discovery (AAAI 2025 · similarity 0.84)
- Finite Resources False Discovery Rate Control on Structured Hypothesis Spaces (NeurIPS 2026 · similarity 0.84)
- Calibration without labels in multiple testing (NeurIPS 2026 · similarity 0.83)
- Multivariate Conformal Selection (ICML 2025 · similarity 0.83)
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.