Description du projet
Des algorithmes d’apprentissage profond pour résoudre les problèmes inverses
Les problèmes inverses utilisent des équations différentielles partielles (EDP) pour établir des liens entre des paramètres inconnus et des données expérimentales. Ils trouvent de nombreuses applications, aussi diverses que l’évaluation de la croissance du cancer, la sécurité des infrastructures civiles et l’amélioration de la production d’énergie géothermique. Néanmoins, les approches d’apprentissage profond (AP) pour les EDP présentent des limites, notamment un manque de fondements théoriques et d’interprétabilité, ce qui entrave leur intégration dans des applications à fort enjeu. Le projet IN-DEEP, financé par le programme MSCA, réunit des doctorants et des scientifiques dans le but de former des étudiants de troisième cycle à l’élaboration, à la mise en œuvre et à l’exploitation d’algorithmes d’apprentissage profond fondés sur les connaissances. Ces algorithmes contribueront à résoudre rapidement et avec précision les problèmes inverses posés par les EDP. IN-DEEP s’est engagé à développer des solveurs avancés reposant sur l’AP pour les EDP présentant un intérêt sociétal et industriel important.
Objectif
IN-DEEP is a European Doctoral Network composed of nine doctoral candidates (DCs) and top scientists with complementary areas of expertise in applied mathematics, artificial intelligence, high-performance computing, and engineering applications. Its main goal is to provide high-level training to the nine DCs in designing, implementing, and using explainable knowledge-driven Deep Learning (DL) algorithms for rapidly and accurately solving inverse problems governed by partial differential equations (PDEs).
Inverse problems in which the unknown parameters are connected to experimental measurements through PDEs cover from medical applications - like cancer growth assessment - to the safety of civil infrastructures, and green geophysical applications such as geothermal energy production. Their application value is measured in human lives and society's well-being, which goes beyond any quantifiable amount of money. This is why equipping a new generation of specialists with highly-demanded skills for the upcoming transition toward safe and robust AI-based technologies is imperative.
Despite the promising results in many applications, DL for PDEs has severe limitations. The most troublesome is its lack of a solid theoretical background and explainability, which prevents potential users from integrating them into high-risk applications.
IN-DEEP aims to remove these constraints to unleash the full potential of DL algorithms for PDEs. We will achieve this by: (a) focusing on emerging applications of DL for PDEs with immense societal and/or industrial value, (b) designing mathematics-infused advanced solvers to address them efficiently, and (c) involving, from the beginning, industrial and technological agents which can monitor, upscale, and exploit this knowledge. On the way, we shall establish the foundations of a better knowledge exchange ecosystem amongst the main academic and industrial actors within Europe, disseminating the results worldwide.
Champ scientifique
- medical and health sciencesclinical medicineoncology
- natural sciencesbiological sciencesecologyecosystems
- natural sciencescomputer and information sciencesartificial intelligencemachine learningdeep learning
- natural sciencesmathematicspure mathematicsmathematical analysisdifferential equationspartial differential equations
- engineering and technologyenvironmental engineeringenergy and fuelsrenewable energygeothermal energy
Mots‑clés
Programme(s)
- HORIZON.1.2 - Marie Skłodowska-Curie Actions (MSCA) Main Programme
Régime de financement
HORIZON-TMA-MSCA-DN - HORIZON TMA MSCA Doctoral NetworksCoordinateur
48940 Leioa
Espagne