Physics contrained machine learning
Description
Objectifs
At the end of this module, the student will have understood and be able to explain (main concepts):
-Main approaches for solving time dependent problem (EDP and Data assimilation) using ML
-Relevance of using physical constraints for solving problems with underling physics (feature engineering), design of Neural networks
-Methods for handling nonlinearity and large scale (use of latent space, high performance computing)
-Performance of ML for solving problems with physical constraints.
At the end of this module, the student should be able to:
-Use ML for solving time dependent PDE and analysis the accuracy
-Analysis the HP performance of the solvers, and propose algorithmic enhancements
-Design a full data assimilation system based on ML, starting from a description of a system using partial differential equation and and observational system
-Assess the performance of a system, question the relevance of the mathematical assumptions
Pré-requis
Numerical algebra for large scale, statistical estimation, non-convex smooth optimization, numerical solution of PDEs, data assimilation, machine learning
Évaluation
L’évaluation des acquis d’apprentissage est réalisée en continu tout le long du semestre. En fonction des enseignements, elle peut prendre différentes formes : examen écrit, oral, compte-rendu, rapport écrit, évaluation par les pairs…
En bref
Crédits ECTS : 3.0
Nombre d’heures : 59.0

INSA Toulouse
135 avenue de Rangueil
31077 Toulouse cedex 4
Tél : 05 61 55 95 13
Fax : 05 61 55 95 00

Dans un souci d'alléger le texte et sans aucune discrimination de genre, l'emploi du genre masculin est utilisé à titre épicène.