EWI building TU Delft

Workshop on Computational and Mathematical Methods in Data Science 2024

Delft University of Technology, April 25-26, 2024

About the Workshop

Welcome to the Workshop on Computational and Mathematical Methods in Data Science 2024. It is a meeting of the Dutch scientific machine learning community, the 2024 edition of the annual workshop of the GAMM Activity Group on “Computational and Mathematical Methods in Data Science” (COMinDS), as well as the first acitivity of the EMS Topical Activity Group Scientific Machine Learning. The workshop is co-organized by the GAMM Activity Group and the Strategic Research Initiative “Bridging Numerical Analysis and Machine Learning” of the 4TU Applied Mathematics Institute (AMI). The workshop will be hosted by Delft University of Technology and take place on April 25 and 26, 2024.

This workshop brings together scientists from mathematics, computer science, and application areas working on computational and mathematical methods in data science.

The meeting will be organized under the support of

Registration

The registration for the GAMM COMinDS Workshop 2024 is closed.

If you have any questions or trouble with the registration, feel free to contact us via email at a.heinlein@tudelft.nl.

Scientific Program

Keynote speakers

Thursday, 25. April 2024

Time (CET) Speaker Topic / Title Room Chair
8.30 - 9.00   Registration Collegezaal  
9.00 - 9.15 Alexander Heinlein Opening & Presentation Strategic Research Initiative Bridging Numerical Analysis and Machine Learning Collegezaal  
9.15 - 10.15 Thomas Richter Hybrid Finite Element / Neural Network simulations (slides) Collegezaal Alexander Heinlein
10.15 - 10.45 Victor Michel-Dansac Approximately well-balanced Discontinuous Galerkin methods using bases enriched with Physics-Informed Neural Networks Collegezaal Alexander Heinlein
10.45 - 11.00   Coffee break Collegezaal  
11.00 - 11.30 Taniya Kapoor Enhancing Generalization in Physics-Informed Machine Learning with Neural Oscillators Collegezaal Janine Weber
11.30 - 12.00 Konrad Janik Neural ODE for Hamiltonian Systems with Irregular and Noisy Data Collegezaal Janine Weber
12.00 - 12.30 Arturo De Marinis Training of stable neural ordinary differential equations (slides) Collegezaal Janine Weber
12.30 - 14.00   Lunch break Collegezaal  
13.00 - 14.00   Poster session Betazaal  
14.00 - 15.00 Victorita Dolean Parallelization approaches for neural network-based collocation methods Collegezaal Thomas Richter
15.00 - 15.30 Felix Dietrich Solving partial differential equations with sampled neural networks Collegezaal Thomas Richter
15.30 - 16.00   Coffee break Collegezaal  
16.00 - 16.30 Hanno Gottschalk Learning to Integrate Collegezaal Axel Klawonn
16.30 - 16.40 Axel Klawonn Presentation EMS Topical Activity Group Scientific Machine Learning Collegezaal  
16.40 - 17.00 Axel Klawonn & Martin Stoll GAMM COMinDS meeting Collegezaal  
17.00 End of the program   Collegezaal  
18.30   Conference dinner:
MOODZ eten & drinken
Vesteplein 4
2611 WG Delft		    

Friday, 26. April 2024

Time (CET) Speaker Topic / Title Room Chair
9.30 - 10.30 Christoph Brune The Data-Weight Duality for Deep Learning Inverse Problems Collegezaal Victorita Dolean
10.30 - 11.00 Jasper Rou Deep Gradient Flow Methods for Option Pricing in Diffusion Models Collegezaal Victorita Dolean
11.00 - 11.30   Coffee break Collegezaal  
11.30 - 12.00 Florian Boßmann Stabilizing neural networks using iterated Graph Laplacian - a seismic impedance example Collegezaal Christoph Brune
12.00 - 12.30 Alexander Litvinenko Computing f-divergences, probability density and characteristic functions in low-rank tensor format (slides) Collegezaal Christoph Brune
12.30 - 13.30   Lunch break Collegezaal  
13.30 - 14.30 Olga Mula Reduced Models in Wasserstein Spaces for Forward and Inverse Problems Collegezaal Martin Stoll
14.30 - 15.00 Francesco Ferranti Response feature-based modeling for small-data machine learning in electromagnetic design Collegezaal Martin Stoll
15.00   Closing with coffee and snacks Collegezaal  

Poster session

Presenter Title
Junaid Akhter Pitfalls to avoid while using multiobjective optimization for machine learning
Amaury Bélières Frendo Geometric shape optimization for Dirichlet energy with physics informed and symplectic neural networks
Roberto Bentivoglio Multi-scale hydraulic-based graph neural networks: generalizing rapid flood mapping to irregular meshes and time-varying boundary condition
Jan Blechschmidt A physics-informed DeepONet model for the solution of quantum graphs
Piero Deidda Nonlinear joint spectral radius of cone order preserving functions
Antoine Lechevallier Hybrid Newton method for the acceleration of well events handling in the numerical simulation of CO2 Storage
Erion Morina On the growth of parameters of approximating neural networks
Moreno Pintore An introduction to the lowest-order Neural Approximated Virtual Element Method
Yong Shang Randomized neural networks with Petrov–Galerkin methods for solving linear elasticity problem
Corné Verburg High-resolution image segmentation with U-Net-based segmentation CNN on multiple GPUs (poster)

Venue & Accommodation

The Workshop on Computational and Mathematical Methods in Data Science will take place in the Collegezaal at the Vakwerkhuis close to the campus of the Delft University of Technology:

      DUWO building
      Professor Schermerhornstraat 4
      2628 PZ Delft
      The Netherlands

More information will follow. Click here for information on how to get to Delft.

For accommodation, we recommend the following hotels:

Hotel Email Address Booking
Hampshire Hotel Delft Centre reservations@hoteldelftcentre.nl Koepoortplaats 3
2612RR Delft		 Booking link
Hotel De Plataan info@hoteldeplataan.nl Doelenplein 10
2611BP Delft		 Discount code:
TUDELFT
thesocialhub.co delft@thesocialhub.co Van Leeuwenhoekpark 1
2611GW Delft		 Discount codes:
TSHPAOLA1 or FORYOU123
Hotel Casa Julia reservations@casajulia.nl Maerten Trompstraat 33
2628RC Delft		 Booking link
Hotel Leeuwenbrug Delft   Koornmarkt 16
2611 EE Delft		 Mention that you are a guest of TU Delft
Hotel Bridges House Delft   Oude Delft 74
2611 CD Delft		 Mention that you are a guest of TU Delft
Hotel Royal Bridges   Koornmarkt 55-65
2611 EC Delft		 Mention that you are a guest of TU Delft
Hotel de Koophandel   Beestenmarkt 30
2611 GC Delft		 Mention that you are a guest of TU Delft

Using the booking links/discount codes in the column Booking, you will be able to book a room at a reduced rate.

Local Organizing Committee