Robert Raymond Stephany

LLNL Logo
  • Email
    stephany1@llnl.gov
  • Phone
    (925) 423-5125
  • Organization
    Not Available

Dr. Robert Stephany is the 2024 Sydney Fernbach Postdoctoral Fellow. His research focuses on applying machine learning to challenges in science and engineering, particularly through the development of algorithms for nonlinear reduced-order modeling (ROM) and the data-driven discovery of partial differential equations. He works on the LaSDI family of ROM algorithms and previously collaborated with Dr. Timo Bremer on the ADMIRRAL project, where he developed novel hierarchical autoencoder architectures for high-dimensional protein data.

Dr. Stephany earned his Bachelor’s degree in Mathematics from the University of Texas at Austin in 2020, followed by a Master’s and Ph.D. in Applied Mathematics from Cornell University in 2024. At Cornell, under the supervision of Dr. Christopher Earls, he developed several algorithms—including PDE-READ, PDE-LEARN, Weak-PDE-LEARN, and DDE-Find—for identifying differential equations from noisy, limited data.

Ph.D. Applied Mathematics, Cornell University, Ithaca, New York

Masters Applied Mathematics, Cornell University, Ithaca, New York

BS Mathematics, University of Texas at Austin, Austin, Texas

Stephany, Robert. "DDE-Find: Learning Delay Differential Equations from Noisy, Limited Data." arXiv preprint arXiv:2405.02661 (2024).
 
Georgouli, Konstantia, Robert R. Stephany, Jeremy OB Tempkin, Claudio Santiago, Fikret Aydin, Mark A. Heimann, Loïc Pottier et al. "Generating Protein Structures for Pathway Discovery Using Deep Learning." Journal of Chemical Theory and Computation 20, no. 20 (2024): 8795-8806.
 
Stephany, Robert, Maria Antonia Oprea, Gabriella Torres Nothaft, Mark Walth, Arnaldo Rodriguez-Gonzalez, and William A. Clark. "Learning the delay in delay differential equations." In ICLR 2024 Workshop on AI4DifferentialEquations In Science. 2024.
 
Stephany, Robert, and Christopher Earls. "Weak-PDE-LEARN: A weak form based approach to discovering PDEs from noisy, limited data." Journal of Computational Physics 506 (2024): 112950.
 
Stephany, Robert, and Christopher Earls. "PDE-LEARN: Using deep learning to discover partial differential equations from noisy, limited data." Neural Networks 174 (2024): 106242.
 
Stephany, Robert, and Christopher Earls. "PDE-READ: Human-readable partial differential equation discovery using deep learning." Neural Networks 154 (2022): 360-382.