Topics
This first worshop will be focused on the use of machine learning techniques in interaction with classical numerical simulations in order to extend their applicability or efficiency. All types of simulations are welcome and a large set of application fields is expected. The workshop will cover in particular the following topics (not restrictive):
- construction of parametric surrogate models from simulation outputs
- model calibration, parameters identification and inverse problems
- construction of hybrid models based on both experimental data and simulation results
- data-based pre-treatments for simulations
- deep learning for geometric design, geometry and mesh generation
- machine learning for post-treatment of simulation outputs and visualization
Additional details will be provided soon …