Cea implementation of machine learning techniques h/f

Le CEA est un acteur majeur de la recherche, au service des citoyens, de l'économie et de l'Etat.

Il apporte des solutions concrètes à leurs besoins dans quatre domaines principaux : transition énergétique, transition numérique, technologies pour la médecine du futur, défense et sécurité sur un socle de recherche fondamentale. Le CEA s'engage depuis plus de 75 ans au service de la souveraineté scientifique, technologique et industrielle de la France et de l'Europe pour un présent et un avenir mieux maîtrisés et plus sûrs.

Implanté au coeur des territoires équipés de très grandes infrastructures de recherche, le CEA dispose d'un large éventail de partenaires académiques et industriels en France, en Europe et à l'international.

Les 20 000 collaboratrices et collaborateurs du CEA partagent trois valeurs fondamentales :

- La conscience des responsabilités
- La coopération
- La curiosité Research Project : Implementation of Machine Learning techniques to optimize laser-plasma accelerators experimentally and numerically for industrial and medical applications.

Since the first experimental evidence for FLASH effect in 2014 with conventional accelerators, the worldwide community is actively working on trying to understand the basic mechanisms underlying the physical and biological process responsible for the increased preservation of healthy tissue surrounding the tumor treated with ultrashort electron bunches. LPAs are promising alternative sources to conventional accelerators for investigating the fundamental physico chemical processes underlying these biological effects.

The first phase of this project will BE to implement a Machine Learning (ML) solution to optimize the properties of the electrons accelerated in our brand new 100 TW-class experimental facility, in order to irradiate various samples (biological, chemical) for dosimetry and radiobiology experiments. This ML solution will leverage the numerous diagnostics that we have already installed in order to BE able to control the properties of the laser and the plasma. As part of the EURO-LABS network (European Laboratories for Accelerator Based
Sciences), we will benefit from a ML toolkit, developed by GSI (Germany). This toolkit, originally conceived for conventional accelerators, will BE adapted to the specificities of LPA and tested on our experimental facility for validation.

The second phase of the project will consist in using ML to implement a surrogate model able to predict the properties of the electrons accelerated with a LPA as a function of the laser and the plasma properties, over a large parameter space. The aim is to use this tool to reduce the need of costly Particle-In-Cell simulations, the numerical tool typically used to model LPAs. The surrogate model will BE trained on numerical simulation results performed with the state-of-the-art Particle-In-Cell code WarpX and, depending on the availability, experimental results.

The third phase of the project would consist in coupling the output of Particle-In-Cell simulations or of the aforementioned surrogate model to a Monte Carlo code such as Geant4-DNA, in order to simulate the effect of electron bunches accelerated with LPAs on biological samples.

Depending on their interest, the postdoc may also BE directly involved in experimental campaigns on the UHI100 laser facility at CEA or on other national and international laser facilities.

This project will benefit from the diverse expertise of the groups at CEA/LIDYL, as well as of their partners. In particular, the second and the third phase of the project will BE carried out in collaboration with the Department of Accelerators, Cryogenics and Magnetism at the Institute for Research on the Fundamental laws of the Universe (CEA/IRFU)