Marta D'Elia
Scholar

Marta D'Elia

Google Scholar ID: h3Q-aHUAAAAJ
Adjunct Professor, Stanford - Director of AI and ModSim, Atomic Machines
Scientific Machine LearningOptimal DesignNonlocal ModelsUncertainty Quantification
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
2,567
 
H-index
28
 
i10-index
51
 
Publications
20
 
Co-authors
11
list available
Contact
LinkedInOpen ↗
Publications
2 items
Physically Interpretable Representation Learning with Gaussian Mixture Variational AutoEncoder (GM-VAE)
2025
Cited
0
Physically Interpretable Representation and Controlled Generation for Turbulence Data
2025
Cited
0
Resume (English only)
Academic Achievements
  • Publications can be followed on Google Scholar.
Research Experience
  • Atomic Machines - Current work involves Scientific Machine Learning, Optimization and Optimal Design, Generative Design, and Digital Twinning in the context of micro-manufacturing.
  • Pasteur Labs - Former Principal Scientist and SciML lead.
  • Meta - Former Research Scientist.
  • Sandia National Laboratories, California - Former Principal Member of the Technical Staff.
  • Stanford - Involved in several SciML projects for scientific applications. Current work involves Graph Neural Nets and Graph-based models, and self-supervised learning for clustering and classification.
Background
  • An expert in AI and ModSim. Currently, the Director of AI and ModSim at Atomic Machines and an Adjunct Professor at Stanford ICME.
Miscellany
  • Can be found on LinkedIn.
Co-authors
11 total
Co-author 1
Co-author 1
Pavel Bochev
Pavel Bochev
Distinguished Member of technical Staff, Sandia National Laboratories
Qiang Du
Qiang Du
Fu Foundation Professor of Applied Mathematics, Columbia University
Co-author 4
Co-author 4
Richard Lehoucq
Richard Lehoucq
Sandia National Labs
Co-author 6
Co-author 6
Eric Phipps
Eric Phipps
Sandia National Laboratories
Co-author 8
Co-author 8