Hi, my name is

Theo Bourdais

PhD Student in Applied and Computational Mathematics

I am a PhD student at Caltech working with Houman Owhadi. I am also grateful to be supported by the Kortschak Scholar program.

I focus on Machine Learning for Scientific Discovery. In my research, I use

  • Gaussian Processes
  • Computer Vision, Natural Language Processing, and Reinforcement Learning
  • Applied Mathematics, such as Random Matrix Theory and Partial Differential Equations

Latest News

03/02/2025

Our paper on Pruning Deep Neural Networks via a Combination of the Marchenko-Pastur Distribution and Regularization was posted on Arxiv!

02/17/2025

I will be giving a talk on Model aggregation at DTE AICOMAS in Paris.

02/01/2025

Our paper on Model aggregation was accepted at ICLR 2025!

02/27/2024

I will be giving a talk on Computational Hypergraph Discovery at the SIAM UQ 2024 conference in Trieste.

Experience

Research Associate - NASA Jet Propulsion Laboratory (JPL)
2024 - 2025
Investigated applying my research to JPL’s aerospace applications, in the context of Digital Twins.
Junior Data Engineer - Doc.ai (ShareCare)
2021 - 2022

Worked on several research projects in Computer Vision for Healthcare:

  • Study of AI-based symptoms tracking for Myasthenia Gravis (collaboration with UCB)
  • Automatic medication label reader development
Research Intern - INRIA
Summer 2020
Research internship on robust optimisation methods and algorithms.

Education

PhD in Applied and Computational Mathematics - California Institute of Technology
2022 - Present

Advisor: Prof. Houman Owhadi. Research focus:

  • Machine learning
  • Gaussian processes
  • Computational mathematics
Master of Mathematics - University of Cambridge
2020 - 2021
Advanced studies in mathematical theory and applications
Master of Science (Engineering) - École Polytechnique
2017 - 2021
Specialized in applied mathematics and computational methods

Research Interests

Algorithm Discovery
Algorithm Discovery

Digital Twins
Digital Twins