Selected Papers

Gradient Methods with Online Scaling
W. Gao, Y. Chu, Y. Ye, and M. Udell
Conference on Learning Theory (COLT), 2025
[arxiv][url][bib]

OptiMUS: Scalable Optimization Modeling Using MIP Solvers and Large Language Models
A. AhmadiTeshnizi, W. Gao, and M. Udell
International Conference on Machine Learning (ICML), 2024
[arxiv][url][bib][video]

Challenges in Training PINNs: A Loss Landscape Perspective
P. Rathore, W. Lei, Z. Frangella, L. Lu, and M. Udell
International Conference on Machine Learning (ICML), 2024
[arxiv][bib]

PROMISE: Preconditioned Stochastic Optimization Methods by Incorporating Scalable Curvature Estimates
P. Rathore, Z. Frangella, and M. Udell
JMLR, 2024
[arxiv][url][bib]

Randomized Numerical Linear Algebra for Optimization
M. Udell and Z. Frangella
SIAG/OPT Views and News, 2022
[pdf][url][bib][video]

Randomized Nystr\"om Preconditioning
Z. Frangella, J. A. Tropp, and M. Udell
SIAM Journal on Matrix Analysis and Applications, 2022
[arxiv][url][bib]

An Optimal-Storage Approach to Semidefinite Programming using Approximate Complementarity
L. Ding, A. Yurtsever, V. Cevher, J. Tropp, and M. Udell
SIAM Journal on Optimization (SIOPT), 2021
Winner of 2017 INFORMS Optimization Society student paper prize
[arxiv][pdf][slides][bib]

Big Data is Low Rank
M. Udell
SIAG/OPT Views and News, 2019
[url][slides][bib]

Why are Big Data Matrices Approximately Low Rank?
M. Udell and A. Townsend
SIAM Mathematics of Data Science (SIMODS), 2019
[arxiv][bib]