Available theses

Emulate a low energy nuclear interaction model with Deep Learning

proponenti: Carlo Mancini Terracciano, Stefano Giagu

gruppo di ricerca: MC-INFN / GeNIALE

titolo della tesi: Emulate a low energy nuclear interaction model with Deep Learning

descrizione: A reliable model to simulate nuclear interactions is fundamental for Ion-therapy. Today, Geant4 is the leading tool to develop Monte Carlo simulation in High Energy Physics, and is widely used in many other application areas, including ion-therapy. These applications are hampered, however, as the physics models currently available in Geant4 fail to reproduce the nuclear fragmentation process in interactions below 100 MeV/u. We already showed how BLOB (“Boltzmann-Langevin One Body”), a model developed to simulate heavy ion interactions up to few hundreds of MeV/u, could simulate also $^{12}$C reactions in the same energy domain. However, its computation time is too long for any medical application. For this reason we are exploring the possibility of emulating it with a Deep Learning algorithm and with generative algorithms in particular. A preliminary paper is under publication and the student will participate in developing new Deep Learning algorithm, will train them with the BLOB output and finally will compare the Deep Learning prediction with BLOB and experimental data.