Ongoing · 2024–present
LiL-Q: convex training for physics-informed models
Recasts physics-informed neural network training for nonlinear PDEs as a short sequence of linear least-squares solves, removing the nonconvexity that makes standard training slow and unreliable. Tested on seven benchmark problems and on reservoir flow, against neural-network and finite-volume baselines.
- Single-digit outer iterations on seven benchmark PDEs
- Runtimes up to three orders of magnitude shorter than standard physics-informed training
- Open-source code with a reproduction guide, archived on Zenodo
Lead developer, M.S. thesis and PhD research
