Research Interests

  • Scientific Machine Learning: Neural Operators, Inverse Problems.
  • Numerical Methods for Partial Differential Equations: finite element methods, iterative solvers.
  • Computational Electromagnetics.
  • Computational Fluid Dynamics. 

EIT prediction

Deep Learning \(\times\) Inverse Problems

Study how to use neural operator to enhance non-iterative methods (such as direct sampling) for inverse problems in the regime of limited data.

We are interested particularly in the Calderón problem.

Neural Operator-Assisted Numerical PDE Solver

Leveraging neural operators' representation capacity to enhance classical numerical schemes, such as finite volume methods or finite element methods. 

Numerical Methods for PDEs

Study of how to use structure-preserving finite dimensional subspaces to approximate partial differential equations.