AI Paper Atlas

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

Predicting Kernel Regression Learning Curves from Only Raw Data Statistics

Dhruva Karkada, Joseph Turnbull, Yuxi Liu, James B Simon

ICLR 2026 · Learning theory and algorithms › Kernels and regression theory · approach: Kernel ridge regression analysis

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
decrease Gaussian kernel width so level coefficients decay more slowly (ablation)HEA for Gaussian kernel · Gaussian data / real image dataagreement with HEA eigenstructureHurt
decrease effective data dimension (ablation)HEA for Laplace kernel and ReLU NTK · real image dataagreement with HEA eigenstructureHurt
decrease effective data dimension (ablation)HEA for Gaussian kernel · real image dataagreement with HEA eigenstructureNo clear effect
use simpler, less Gaussian datasets (CIFAR -> SVHN -> MNIST -> tabular) (ablation)HEA · kernel eigenstructure on image/tabular datasetsagreement with HEA eigenstructureHurt
re-orthogonalize sampled Hermite polynomials with Gram-Schmidt when estimating target coefficientsHEA learning-curve prediction · KRR learning curves on real datasetstarget coefficient estimationImproved
truncate the HEA at finite order (5 to 10) and use high effective dimensionHEA for Laplace and ReLU kernels · real image dataagreement with HEA eigenstructureImproved
apply sample normalization or ZCA whitening to the dataHEA · image datasetsagreement with HEA eigenstructureImproved
predict KRR learning curves with HEA eigenstructure plus eigenframeworkKRR with Gaussian, Laplace, ReLU NNGP, NTK kernels · CIFAR-5m, SVHN, ImageNettest risk vs sample sizeImproved
use HEA eigenvalue to predict MLP training time for Hermite targetsthree-layer ReLU MLP in feature-learning regime · Gaussian data and CIFAR-5m multi-Hermite targetssteps to reach online MSE <= 0.1Improved

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