Mathematical data science

Michael R. Douglas, Kyu-Hwan LeeAdvances in Theoretical and Mathematical Physics 2026 · DOI · arXiv

Abstract

This article explores how machine learning, through a paradigm of mathematical data science, can enable the discovery of new mathematical structures by analyzing datasets of mathematical objects, illustrated by case studies on murmurations in number theory and partition loadings related to Kronecker coefficients.
BibTeX
@inproceedings{douglas2026mathematical,
  author = {Michael R. Douglas and Kyu-Hwan Lee},
  title = {{Mathematical data science}},
  booktitle = {Advances in Theoretical and Mathematical Physics},
  year = {2026},
  doi = {10.4310/atmp.260412174613},
}

Loading related content...

Figure