Teaching

All at École nationale des ponts et chaussées.

Courses

  1. Feb – Apr 2026

    Practice of machine learning methods (PRAMA) 12 h

    Practical sessions in Python: random forests, gradient boosting, support vector machines and neural networks applied to real datasets.

  2. Nov 2025 – Jan 2026

    Analysis and PDEs, part I 20 h

    First-year course, lectures and exercise sessions: normed vector spaces, Banach and Hilbert spaces, measure theory and Lebesgue integration.

  3. Nov 2024 – Jan 2025

    Analysis and PDEs, part II 20 h

    First-year course, lectures and exercise sessions: Lebesgue spaces, Fourier series, distribution theory, Fourier transform.

Student projects

First-year research projects that I proposed and supervised.

  1. Feb – May 2026

    Neural networks for solving partial differential equations 12 h

    Students implemented neural networks in Python, used them to solve 1D and 2D PDEs with physics-informed neural networks (PINNs) and the Deep Ritz method, and compared the results with classical methods.

  2. Feb – May 2026

    Model order reduction: POD and hyper-reduction (DEIM) 12 h

    Finite element solution of simple parametric PDEs (variational formulation, Lax–Milgram, Galerkin method), then hyper-reduced reduced-order models with POD and DEIM, implemented in Python.