Research map / Learning theory and algorithms / Kernels and regression theory
Can we globally optimize cross-validation loss? Quasiconvexity in ridge regression
William T. Stephenson, Zachary Frangella, Madeleine Udell, Tamara Broderick
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 PCR preprocessing with R = 15 | Ridge regression LOOCV on life-expectancy data · Life expectancy | LOOCV quasiconvexity | Hurt |
| Use a subset of 50 observations | Ridge regression LOOCV on wine-quality data · Wine quality | LOOCV quasiconvexity | Hurt |
| Make the covariate spectrum sufficiently close to uniform, under the theorem's assumptions | Ridge regression LOOCV · Theoretical setting with sufficiently large N | LOOCV quasiconvexity | Improved |
| Use i.i.d. sub-Gaussian covariates in a well-specified linear model with sufficiently small noise and a consistent least-squares estimator | Ridge regression LOOCV · Well-specified linear model | LOOCV quasiconvexity | Improved |
| Change the right singular vectors, response norm, or overall singular-value scale | Ridge regression LOOCV · General data under the paper's preprocessing assumptions | LOOCV quasiconvexity | No clear effect |
| Violate the assumed decay of the largest row norm of the left singular-vector matrix | Ridge regression LOOCV with a uniform spectrum · Simulated data | LOOCV quasiconvexity | Hurt |
| Increase the least-squares residual norm beyond the theorem's assumed range | Ridge regression LOOCV · Simulated data | LOOCV quasiconvexity | Mixed |
Closest papers by meaning
- Risk and cross validation in ridge regression with correlated samples (ICML 2025 · similarity 0.88)
- An Agnostic View on the Cost of Overfitting in (Kernel) Ridge Regression (ICLR 2024 · similarity 0.85)
- Asymptotically Free Sketched Ridge Ensembles: Risks, Cross-Validation, and Tuning (ICLR 2024 · similarity 0.84)
- Is Cross-validation the Gold Standard to Estimate Out-of-sample Model Performance? (NeurIPS 2024 · similarity 0.83)
- Optimal Ridge Regularization for Out-of-Distribution Prediction (ICML 2024 · similarity 0.83)
- Regression Learning with Limited Observations of Multivariate Outcomes and Features (ICML 2024 · 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.
