Generalization error rates in kernel regression: the crossover from the noiseless to noisy regime*

DOI10.1088/1742-5468/AC9829OpenAlexW3213347999MaRDI QIDQ5055412FDOQ5055412

Authors: Hugo Cui, Bruno Loureiro, Florent Krzakala, Lenka Zdeborová 

Publication date: 13 December 2022

Published in: Journal of Statistical Mechanics: Theory and Experiment (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2105.15004

Recommendations

• Just interpolate: kernel ``ridgeless regression can generalize
• Kernel regression, minimax rates and effective dimensionality: beyond the regular case
• Regularization in kernel learning
• Generalization bounds of a compressed regression learning algorithm
• Learning with generalization capability by kernel methods of bounded complexity

zbMATH Keywords

learning theorymachine learning

Mathematics Subject Classification ID

Statistical mechanics, structure of matter (82-XX)

Cites Work

• Scikit-learn: machine learning in Python
• Title not available (Why is that?)
• Title not available (Why is that?)
• Acceleration of Stochastic Approximation by Averaging
• Title not available (Why is that?)
• Eigenvectors of some large sample covariance matrix ensembles
• Optimal rates for the regularized least-squares algorithm
• Concentration of measure and isoperimetric inequalities in product spaces
• Bayesian learning for neural networks
• Ridge regression and asymptotic minimax estimation over spheres of growing dimension
• Support vector machines learning noisy polynomial rules
• Sobolev norm learning rates for regularized least-squares algorithms
• Precise Error Analysis of Regularized <inline-formula> <tex-math notation="LaTeX">$M$ </tex-math> </inline-formula>-Estimators in High Dimensions
• Optimal rates for spectral algorithms with least-squares regression over Hilbert spaces
• High-dimensional dynamics of generalization error in neural networks
• High-dimensional asymptotics of prediction: ridge regression and classification
• Benign overfitting in linear regression
• Title not available (Why is that?)
• Surprises in high-dimensional ridgeless least squares interpolation
• Two models of double descent for weak features
• When do neural networks outperform kernel methods?*
• Asymptotic learning curves of kernel methods: empirical data versus teacher–student paradigm

Cited In (2)

• Learning sparse features can lead to overfitting in neural networks
• On expected error of randomized Nyström kernel regression

Uses Software

• Scikit

This page was built for publication: Generalization error rates in kernel regression: the crossover from the noiseless to noisy regime*

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5055412)