Research map / Learning theory and algorithms / Prediction and decision losses
The knockoff filter for FDR control in group-sparse and multitask regression
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 group-level knockoff construction and group-lasso filtering instead of feature-level knockoffs | Group knockoff · Group-sparse simulations | power | Helped |
| Use group-level rather than feature-level knockoffs | Group knockoff · Group-sparse simulations with increasing within-group correlation | power | Helped |
| Add the conservative plus correction to the group knockoff filter (ablation) | Group knockoff · Group-sparse simulations | power | Hurt |
| Add the conservative plus correction to the knockoff filter (ablation) | Knockoff · Group-sparse simulations | power | Hurt |
| Use the group knockoff filter | Group knockoff · Group-sparse simulations | group FDR control | No clear effect |
| Use multitask knockoff instead of pooled knockoff | Multitask regression · Multitask simulations | power | Helped |
| Use parallel knockoff, combining discoveries from separate response regressions | Multitask regression · Multitask simulations | FDR control | Hurt |
| Use parallel knockoff instead of multitask or pooled knockoff | Multitask regression · Multitask simulations | power | Helped |
| Use multitask knockoff or pooled knockoff | Multitask regression · Multitask simulations | FDR control | Mixed |
| Use multitask knockoff instead of parallel knockoff | HIV-1 drug resistance analysis · PI, NRTI, and NNRTI drug classes | agreement with TSM panel | Helped |
| Use multitask knockoff instead of parallel knockoff | HIV-1 drug resistance analysis · PI, NRTI, and NNRTI drug classes | number of discoveries | Hurt |
Closest papers by meaning
- Split Group Knockoffs: Controlling False Discovery Rate in Transformational Group Sparsity (ICML 2026 · similarity 0.88)
- False Discovery Proportion control for aggregated Knockoffs (NeurIPS 2023 · similarity 0.86)
- Dual Feature Reduction for the Sparse-group Lasso and its Adaptive Variant (ICML 2025 · similarity 0.84)
- Knockoffs Inference for Partially Linear Models with Automatic Structure Discovery (AAAI 2025 · similarity 0.84)
- Model-Agnostic FDR Control via Group Gaussian Mirror and Permutation SHAP (NeurIPS 2026 · similarity 0.84)
- Self-Organized Conformal Prediction: Reducing Regional Coverage Gaps with Unsupervised Group Discovery (NeurIPS 2026 · 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.