Research map / Learning theory and algorithms / Prediction and decision losses
More Powerful and General Selective Inference for Stepwise Feature Selection using Homotopy Method
Kazuya Sugiyama, Vo Nguyen Le Duy, Ichiro Takeuchi
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 |
|---|---|---|---|
| Minimal conditioning (Homotopy: condition only on selected features) instead of additionally conditioning on history H and signs S (Polytope) | Forward SFS selective inference · Synthetic data, TPR experiment | true positive rate | Improved |
| Additionally conditioning on history H and/or signs S (ablation) | Forward SFS selective inference · Synthetic data, TPR experiment | true positive rate | Hurt |
| Minimal conditioning (Homotopy) instead of Polytope conditioning | Forward SFS selective inference · Real data (Housing, Abalone, Concrete), subsample sizes 25, 50, 100 | confidence interval length | Improved |
| Minimal conditioning (Homotopy) instead of quadratic conditioning on history (Quadratic) | FB-SFS with AIC stopping selective inference · Synthetic data, TPR experiment | true positive rate | Improved |
| Minimal conditioning (Homotopy) instead of quadratic conditioning on history (Quadratic) | FB-SFS with AIC stopping selective inference · Real data (Housing, Abalone, Concrete), subsample sizes 25, 50, 100 | confidence interval length | Improved |
| Homotopy-based conditional SI (minimal conditioning) for forward SFS and FB-SFS | Forward SFS and FB-SFS selective inference · Synthetic data, FPR experiments | false positive rate | No clear effect |
Closest papers by meaning
- More Powerful Conditional Selective Inference for Generalized Lasso by Parametric Programming (JMLR 2022 · similarity 0.89)
- Statistical Test for Feature Selection Pipelines by Selective Inference (ICML 2025 · similarity 0.85)
- Estimating Conditional Mutual Information for Dynamic Feature Selection (ICLR 2024 · similarity 0.85)
- Kernel Feature Selection via Conditional Covariance Minimization (NeurIPS 2017 · similarity 0.85)
- Contextual Feature Selection with Conditional Stochastic Gates (ICML 2024 · similarity 0.85)
- SAND: One-Shot Feature Selection with Additive Noise Distortion (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.
