BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//AIP Calendar//www.aip.de//
BEGIN:VEVENT
SUMMARY:Colloquium | Tobias Buck (Heidelberg)
DTSTART:20261119T133000Z
DTEND:20261119T143000Z
DTSTAMP:20261115T185110Z
UID:www.aip.de-2365
CATEGORIES:Wissenschaftliches
CATEGORIES:Wissenschaftlicher Vortrag
CATEGORIES:öffentlich
CATEGORIES:Kolloquium
CONTACT:Ewald Puchwein <epuchwein@aip.de>
DESCRIPTION:Speaker: Tobias Buck (Heidelberg)\nTitle: Machine Learning in 
 Astrophysics: From Black-Box Magic to New Scientific Methods\nAbstract: Ma
 chine learning has moved from the margins of astrophysics into the core of
  how we model\, infer\, and explore the Universe. Yet to many it still loo
 ks like black magic: powerful\, opaque\, and hard to trust. In this talk I
  argue that\, used carefully\, ML is becoming a genuine scientific method\
 , one that extends rather than replaces the classical interplay of theory\
 , simulation\, and observation. I organise the discussion around three que
 stions we ask of any physical model. \nCan we predict? Neural emulators re
 place expensive forward models\, from galactic chemical evolution to stiff
  astrochemical networks compressed into low-dimensional latent dynamics wi
 th neural ODEs. GPU-native\, end-to-end differentiable simulations in JAX\
 , spanning N-body dynamics and multiphase galactic winds\, turn the simula
 tor itself into an object we can optimise and learn. \nCan we infer? Simul
 ation-based inference provides principled inverse modelling\, for example 
 recovering the accretion history of the Milky Way from stellar dynamics an
 d chemical abundances. Differentiable simulators open the door to gradient
 -based inference directly through the physics. \nCan we see structure? Clu
 stering and representation learning reveal substructure in high-dimensiona
 l data\, from accreted debris in phase space to galaxies and their haloes.
  \nThroughout\, I emphasise what separates method from magic: calibrated u
 ncertainties\, validation against physics\, and open\, reproducible softwa
 re.\n
LOCATION:AIP\; Kirch-Haus\, Konferenzraum
END:VEVENT
END:VCALENDAR