Learning (cs.LG)

Online Learning to Transport via the Minimal Selection Principle

Universal Prediction Band via Semi-Definite Programming

A Precise High-Dimensional Asymptotic Theory for Boosting and Minimum-L1-Norm Interpolated Classifiers

Reversible Gromov-Monge Sampler for Simulation-Based Inference

How Well Generative Adversarial Networks Learn Distributions

Mehler’s Formula, Branching Process, and Compositional Kernels of Deep Neural Networks

Interpolating Classifiers Make Few Mistakes

Deep Neural Networks for Estimation and Inference

Training Neural Networks as Learning Data-adaptive Kernels: Provable Representation and Approximation Benefits

On the Multiple Descent of Minimum-Norm Interpolants and Restricted Lower Isometry of Kernels