About the Workshop

High-dimensional learning and inference have recently seen a string of rigorous results showing that both learning performance and dynamics obey universal mean-field laws.

This workshop aims to bring together researchers from information theory, machine learning (ML), high-dimensional statistics, random matrix theory, and statistical physics, to develop a unified view of these advances. Topics include dynamical mean-field limits for learning algorithms, universality and Gaussian equivalence in and beyond generalized linear models, universality of iterative algorithms such as AMP, and information-theoretic questions motivated by these asymptotic laws.

Organizers

Rishabh Dudeja UW Madison

Zhenyu Liao HUST

Junjie Ma CAS

Arian Maleki Columbia University

Contact

Page views: | Visitors: