· Mateo Díaz

Mateo Díaz #

Postdoctoral Scholar
Computing and Mathematical Sciences
Caltech

About #

I am Postdoctoral Scholar at Caltech hosted by Venkat Chandrasekaran and Joel Tropp. I obtained my PhD in Applied Mathematics from Cornell University advised by Damek Davis. Before Cornell, I completed a MSc in Mathematics and two BS in Mathematics, and Systems and Computing Engineering at Universidad de los Andes. There I was co-advised by Mauricio Junca and Mauricio Velasco. I spent the Fall of 2020 with the Algorithms and Optimization team at Google Research, hosted by Miles Lubin and David Applegate.

Curriculum vitae

Research interests #

My research interests at the beautiful interplay between continuous optimization, geometry, and statistics and its applications to data science, machine learning and signal processing.

Contact #

email: < first_name > dd < at > caltech < dot > edu
office: Annenberg 305
mail: 1200 E. California Blvd., Mail Code 305-16, Pasadena, CA 91125

Publications #

Clustering a Mixture of Gaussians with Unknown Covariance
(with D. Davis and K. Wang) Submitted, 2021.
Escaping strict saddle points of the Moreau envelope in nonsmooth optimization
(with D. Davis and D. Drusvyatskiy) Submitted, 2021.
Optimal Convergence Rates for the Proximal Bundle Method
(with B. Grimmer) Submitted, 2021.
Infeasibility detection with primal-dual hybrid gradient for large-scale linear programming
(with D. Applegate, H. Lu, and M. Lubin) Submitted, 2021.
Practical Large-Scale Linear Programming using Primal-Dual Hybrid Gradient
(with D. Applegate, O. Hinder, H. Lu, M. Lubin, B. O’Donoghue, and W. Schudy) NeurIPS, 2021.
Low-rank matrix recovery with composite optimization: good conditioning and rapid convergence
(with V. Charisopoulos, Y. Chen, D. Davis, L. Ding, D. Drusvyatskiy) Foundations of Computational Mathematics, 2021.
Efficient Clustering for Stretched Mixtures: Landscape and Optimality
(with K. Wang and Y. Yan) NeurIPS, 2020.
Composite optimization for robust rank one bilinear sensing
(with V. Charisopoulos, D. Davis, and D. Drusvyatskiy) Information and Inference, 2020.
Local angles and dimension estimation from data on manifolds
(with A. Quiroz, M. Velasco) Journal of Multivariate Analysis, 2019.
Compressed sensing of data with known distribution
(with M. Junca, F. Rincón and M. Velasco) Applied and Computational Harmonic Analysis, 2018.
In Search of Balance: The Challenge of Generating Balanced Latin Rectangles
(with C. Gomes, R. Le Bras) CPAIOR 2017.