Talks

2026

Solving Sharp-Interface Phase-Change Problems: Newton Is All You Need!

MSME Seminar, Université Gustave Eiffel, France

May 2026

2026

AI Tools for Research - Feedback after 2 years

MSME Seminar, Université Gustave Eiffel, France

Avril 2026

2025

A space-time cut-cell method for phase change: from stationary geometries to moving interfaces

MSME Seminar, Université Gustave Eiffel, France

Avril 2025

2024

Solving Partial Differential Equations using artificial neural networks

MSME Seminar, Université Gustave Eiffel, France

February 2024

This talk explores the use of artificial neural networks for solving partial differential equations (PDEs) without traditional mesh-based discretization, by reformulating the problem as an optimization over network parameters. The Evolutional Deep Neural Network (EDNN) approach is demonstrated on Ginzburg–Landau and Taylor–Green benchmarks, with discussions on network architecture, loss landscapes, computational cost, and current limitations.