Sharp Generalization for Nonparametric Regression in Interpolation Space by Over-Parameterized Neural Networks Trained with Preconditioned Gradient Descent and Early Stopping

arXiv:2407.11353 · stat.ML, cs.LG, math.ST, stat.TH · Submitted 2024-07-16 · Read on arXiv

stat.ML, cs.LG, math.ST, stat.TH

Submitted: 2024-07-16

Updated: 2026-10-26

Terminology

• Generalization — 2× in this paper · explained in 14 episodes
• Preconditioned Gradient Descent — 2× in this paper
• Hilbert Space — 1× in this paper · explained in 2 episodes
• Neural Tangent Kernel — 1× in this paper
• Reproducing Kernel Hilbert Space (RKHS) — 1× in this paper · explained in 7 episodes

Related papers

• Behavior of prediction performance metrics with rare events
• Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
• A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
• Scalable Posterior Uncertainty for Flexible Density-Based Clustering
• A Transport Geometry of Benign Overfitting
• Efficient quadratic entropy with distance sketches