Séminaire Doctoral / Seminar PhD |
« LyAl-Net: A high-efficiency Lyman-a forest simulation with neural network » |
Naï Boonkongkird |
The inference of cosmological quantities requires accurate and large cosmological simulations. However, their computational time can take millions of CPU hours for a modest coverage in cosmological scales (~(100 Mpc/h)^3). In this machine learning approach, we have trained the U-net with the Horizon-NoAGN simulation to predict the neutral hydrogen physical properties; density, temperature, and line-of-sight velocity fields from dark matter density field. This approach provides extremely fast and robust numerical simulations of Lyman-a forest with a resolution of R ? 30000. |
vendredi 26 novembre 2021 - 16:00 Salle du Conseil, Institut d'Astrophysique |
Page web du séminaire / Seminar's webpage |