Research map / Learning theory and algorithms / Sample complexity
Near-Optimal Linear Regression under Distribution Shift
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 |
|---|---|---|---|
| Adapt regularization to source and target input covariances | Linear regression under covariate shift · Synthetic covariate-shift experiments | target excess risk | Helped |
| Use target-covariance-based minimax shrinkage instead of ordinary ridge shrinkage | Linear regression under covariate shift · Synthetic covariate-shift experiments | target excess risk: \mathcal{O}(d^{-1/4}) | Helped |
| Use source-only ridge regression despite model shift | Linear regression with model shift · Synthetic model-shift experiments | target excess risk | Hurt |
| Use target-domain labeled data for ridge regression rather than source-only ridge | Linear regression with model shift · Synthetic model-shift experiments | target excess risk | Helped |
| Combine source and target information with the proposed minimax estimator | Linear regression with model shift · Synthetic model-shift experiments | target excess risk | Helped |
| Increase model shift | Proposed minimax estimator · Synthetic model-shift experiments | target excess risk | Hurt |
| Apply importance reweighting | Weighted least squares · Synthetic approximation-error experiment | target excess risk | Helped |
| Use importance reweighting without variance-reducing minimax shrinkage | Weighted least squares · Synthetic approximation-error experiment | target excess risk | Hurt |
| Reweight, then apply minimax variance reduction | Approximation-error estimator · Synthetic approximation-error experiment | target excess risk | Helped |
Closest papers by meaning
- Minimum-Norm Interpolation Under Covariate Shift (ICML 2024 · similarity 0.87)
- Transfer Learning for Benign Overfitting in High-Dimensional Linear Regression (NeurIPS 2025 · similarity 0.86)
- Transfer Learning Meets Functional Linear Regression: No Negative Transfer Under Posterior Drift (AAAI 2025 · similarity 0.86)
- Transfer Learning of Linear Regression with Multiple Pretrained Models: Benefiting from More Pretrained Models via Overparameterization Debiasing (NeurIPS 2026 · similarity 0.86)
- Maximum Likelihood Estimation is All You Need for Well-Specified Covariate Shift (ICLR 2024 · similarity 0.85)
- Tolerant Algorithms for Learning with Arbitrary Covariate Shift (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.