AI Paper Atlas

Research map / Learning theory and algorithms / Kernels and regression theory

On Optimal Interpolation in Linear Regression

Eduard Oravkin, Patrick Rebeschini

NeurIPS 2021 · Learning theory and algorithms › Kernels and regression theory · approach: High-dimensional linear regression theory · 0 citations (OpenAlex)

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 changedWhereMeasuredEffect
Use the optimal response-linear achievable interpolator instead of the minimum-norm interpolatorOverparameterized linear regression · Strong and weak features covariance model, with γ = 2 and ψ₁ = 1/2Asymptotic excess riskImproved
Use the optimal response-linear achievable interpolator instead of the optimal-variance interpolatorOverparameterized linear regression · Strong and weak features covariance model, with γ = 2 and ψ₁ = 1/2Asymptotic excess riskImproved
Approximate the population covariance with Graphical Lasso and estimate the signal-to-noise ratio by cross-validationEmpirical optimal response-linear interpolator · Strong and weak features covariance modelGeneralization errorNo clear effect
Use the empirical interpolator without prior-covariance information instead of the optimal interpolatorEmpirical optimal response-linear interpolator · Non-isotropic prior regimesGeneralization performanceNo clear effect
Replace population-covariance preconditioning with the empirical covariance XᵀX/n + λIPreconditioned gradient descent · Overparameterized linear regressionGeneralizationHurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy