Research map / Learning theory and algorithms / Sample complexity
Estimating the Lasso's Effective Noise
Johannes Lederer, Michael Vogt
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 bootstrap-estimated quantile lambda_hat instead of cross-validated lambda for lasso tuning | lasso · Variable selection (Hamming distance, kappa=0.25) | Hamming distance | Improved |
| use bootstrap-estimated quantile instead of cross-validation | lasso · l1-distance | l1 error | Improved |
| use bootstrap-estimated quantile instead of cross-validation | lasso · l_infinity distance | l_inf error | Hurt |
| use bootstrap-estimated quantile instead of cross-validation | lasso · prediction error | prediction error | Hurt |
| use estimated quantile lambda_hat instead of oracle quantile lambda* | lasso tuning · Hamming, l1, l_inf, prediction error | all four errors | No clear effect |
| lower covariate correlation kappa from 0.25 to 0 | lasso with estimated lambda · Variable selection | Hamming distance | Improved |
| replace oracle test by feasible bootstrap test at alpha=0.05 | H0: beta*=0 test · size, n=500, p=250 | empirical size: 0.056 -> 0.057 | No clear effect |
| replace oracle test by feasible bootstrap test at alpha=0.05 | H0: beta*=0 test · power SNR=0.2, n=500, p=1000 | empirical power: 0.867 -> 0.811 | No clear effect |
| increase SNR from 0.1 to 0.2 | feasible test · power, n=500, p=250, alpha=0.05 | empirical power: 0.304 -> 0.850 | Improved |
| increase dimension p from 250 to 1000 | feasible test · power SNR=0.1, alpha=0.05 | empirical power: 0.304 -> 0.284 | Hurt |
Closest papers by meaning
- Handling correlated and repeated measurements with the smoothed multivariate square-root Lasso (NeurIPS 2019 · similarity 0.88)
- A Practical Scheme and Fast Algorithm to Tune the Lasso With Optimality Guarantees (JMLR 2016 · similarity 0.87)
- The Complete Lasso Tradeoff Diagram (NeurIPS 2020 · similarity 0.86)
- Price of Quality: Sufficient Conditions for Sparse Recovery using Mixed-Quality Data (ICLR 2026 · similarity 0.85)
- Non-Asymptotic Uncertainty Quantification in High-Dimensional Learning (NeurIPS 2024 · similarity 0.85)
- Robust Sparse Regression with Non-Isotropic Designs (NeurIPS 2024 · similarity 0.84)
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.
