\[ \begin{align*} \nabla \cdot \mathbf{E} &= 0 \\ \nabla \cdot \mathbf{B} &= 0 \\ \nabla \times \mathbf{E} &= -\frac{\partial \mathbf{B}}{\partial t} \\ \nabla \times \mathbf{B} &= \frac{\partial \mathbf{E}}{\partial t} \end{align*} \]

Modeling Linear PDE Systems with Probabilistic Machine Learning

Bogdan Raiță
Department of Mathematics and Statistics
Georgetown University

With many, many contributions by Marc Härkönen, Markus Lange-Hegermann, Jianlei Huang, and Xin Li

Our goals

  • Use additional knowledge of linear PDE systems with constant coefficients.
  • Construct a computationally nice probabily distribution on such PDE systems.
  • Restricting to this important special case allows more suitable methods.
  • Introduce two such methods: EPGP and S-EPGP.