Research map / Learning theory and algorithms / Matrix and tensor methods
Experimental Design on a Budget for Sparse Linear Models and Applications
Sathya Narayanan Ravi, Vamsi Ithapu, Sterling Johnson, Vikas Singh
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 ED-I or ED-S instead of baseline designs (including 1-mean) for choosing subjects | Lasso on reduced subject set · prostate and lars datasets | feature selection error vs full Lasso | Improved |
| Increase budget (subjects selected) (ablation) | ED-I / ED-S · neuro dataset | zero inconsistency | Improved |
| Increase number of allowed Lasso nonzero features (ablation) | feature selection error of ED-I/ED-S · prostate and lars datasets | selection error | Hurt |
| Reduced (budgeted) setup versus full model | ED-I / ED-S · neuro dataset | BIC | Improved |
| Reduced (budgeted) setup versus full model | ED-I / ED-S · neuro dataset | adjusted R2 | Improved |
| Increase budget (ablation) | ED-I / ED-S · neuro dataset | R2 gain over full model | Hurt |
| Small budget with many allowed nonzero Lasso features (ablation) | ED-I / ED-S · neuro dataset | sign inconsistency | Hurt |
| Use ED-S instead of ED-I | budgeted design · neuro dataset | model-fit measures | No clear effect |
| Use ED-S instead of ED-I | budgeted design · neuro dataset | hypothesis test rejections vs full model | Hurt |
| Reduced vs full model sample moments | ED-I / ED-S · neuro dataset | ratio of moments 1-4 | No clear effect |
Closest papers by meaning
- Fast Screening Rules for Optimal Design via Quadratic Lasso Reformulation (JMLR 2023 · similarity 0.90)
- On Computationally Tractable Selection of Experiments in Measurement-Constrained Regression Models (JMLR 2017 · similarity 0.88)
- Experimental Design for Linear Functionals in Reproducing Kernel Hilbert Spaces (NeurIPS 2022 · similarity 0.87)
- Feature Adaptation for Sparse Linear Regression (NeurIPS 2023 · similarity 0.87)
- Robust Sparse Regression with Non-Isotropic Designs (NeurIPS 2024 · similarity 0.87)
- Efficient Sublinear-Regret Algorithms for Online Sparse Linear Regression with Limited Observation (NeurIPS 2017 · 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.
