BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250822T115808Z
LOCATION:Room 5.0B56
DTSTART;TZID=Europe/Stockholm:20250616T125000
DTEND;TZID=Europe/Stockholm:20250616T132000
UID:submissions.pasc-conference.org_PASC25_sess109_msa168@linklings.com
SUMMARY:Structure-Preserving Neural Networks for Hamiltonian Systems
DESCRIPTION:Benedikt Brantner and Michael Kraus (Max Planck Institute for 
 Plasma Physics)\n\nIn this talk we perform structure-preserving reduced or
 der modeling for the semi-discretized Hamiltonian PDEs. Reduced order mode
 ling can alleviate the cost involved in applications such as optimization,
  uncertainty quantification and inverse problems, that require the repeate
 d solution of large-dimensional physical systems. For this task we can use
  neural networks among other techniques.\nWe start by giving a short overv
 iew of methods for performing reduced order modeling and how to make them 
 structure-preserving. A focus will be put on symplectic autoencoders. Thes
 e are neural networks that respect the symplectic structure of the underly
 ing differential equation. We will show the advantages of both structure p
 reservation and neural networks, compared to other techniques, when perfor
 ming reduced order modeling for Hamiltonian systems and discuss future cha
 llenges ahead for dealing with real-world magnetic fusion plasmas.\n\nDoma
 in: Physics, Computational Methods and Applied Mathematics\n\nSession Chai
 r: Laurent Villard (EPFL)\n\n
END:VEVENT
END:VCALENDAR