Research map / Learning theory and algorithms / Prediction and decision losses
A General Framework for Sparsity Regularized Feature Selection via Iteratively Reweighted Least Square Minimization
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 small r (<=1) in l2,r loss instead of least squares (r=2) when labels contain outliers (ablation) | GSR-FS · Synthetic data with 1% and 10% outliers | RMSE | Helped |
| use small r (<=1) in l2,r loss instead of least squares (r=2) with no outliers (ablation) | GSR-FS · Synthetic data without outliers | RMSE | Hurt |
| decrease sparsity norm p from 1.0 to 0.75 (ablation) | GSR-FS · ISOLET | accuracy (%): 92.98 -> 93.92 | Helped |
| decrease sparsity norm p from 1.0 to 0.5 (ablation) | GSR-FS · AR | accuracy (%): 92.38 -> 94.51 | Helped |
| decrease sparsity norm p from 0.5 to 0.25 (ablation) | GSR-FS · ORL | accuracy (%): 91.13 -> 90.06 | Hurt |
| decrease sparsity norm p from 0.75 to 0.25 (ablation) | GSR-FS · SEMEION | accuracy (%): 86.97 -> 86.58 | No clear effect |
| replace RK1U (r=2 least squares, p=0.5) with GSR-FS (tuned r, p=0.5) | Sparse feature selection · AR | accuracy (%): 87.90 -> 94.51 | Helped |
| replace RK1U with GSR-FS (p=0.5) on running time | Sparse feature selection · ISOLET | running time (s): 6467.20 -> 21.47 | Helped |
| use sparsity regularized methods instead of filter methods (ReliefF, mRMR) | feature selection · ISOLET | accuracy (%): 85.10 -> 94.10 | Helped |
Closest papers by meaning
- A Probabilistic Derivation of LASSO and L12-Norm Feature Selections (AAAI 2019 · similarity 0.88)
- On Benefits of Selection Diversity via Bilevel Exclusive Sparsity (CVPR 2016 · similarity 0.85)
- Best Subset Selection: Optimal Pursuit for Feature Selection and Elimination (ICML 2025 · similarity 0.84)
- Adaptive Feature Selection: Computationally Efficient Online Sparse Linear Regression under RIP (ICML 2017 · similarity 0.84)
- Uncorrelated Group LASSO (AAAI 2016 · similarity 0.84)
- A Unified Framework for Generalization Error Analysis of Learning with Arbitrary Discrete Weak Features (ICML 2025 · similarity 0.82)
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.