BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Workshop "Multifidelity Methods for Stochastic and Uncertain Probl
 ems" [HIM Workshop]
DTSTART:20260202T080000Z
DTEND:20260206T150000Z
DTSTAMP:20260421T225200Z
UID:indico-event-280@math-events.uni-bonn.de
CONTACT:him-coordination@hcm.uni-bonn.de\;him-contact@hcm.uni-bonn.de
DESCRIPTION:Speakers: Kanami Ueda\, Emma Seggewiss (HCM)\n\nThis workshop 
 is aimed at the participants in the Trimester Program "Computational multi
 fidelity\, multilevel\, and multiscale methods". It is not possible to app
 ly only for this workshop. \nGroup leaders: \n\nEmil Løvbak (Karlsruhe 
 Institute of Technology)\nSteffen W. R. Werner (Virginia Polytechnic Insti
 tute and State University)\n\nExternal Speakers: \n\nSebastian Krumscheid
  (Karlsruhe Institute of Technology)\nGiovanni Samaey (KU Leuven)\nElisabe
 th Ullmann (Technical University of Munich)\nAretha Teckentrup (University
  of Edinburgh)\n\nDescription: \nMany simulation problems contain phenome
 na at multiple scales of resolution in terms of both space and time. In pr
 actice\, solving such problems is often challenging due to inherently stoc
 hastic model behavior or uncertain parameters\, modeled probabilistically.
  As such\, the demand for sophisticated techniques for stochastic simulati
 on and uncertainty quantification (UQ) is growing across scientific and en
 gineering disciplines. Hierarchical methods\, which use multilevel and mul
 tifidelity approaches\, are particularly promising\, balancing accuracy an
 d computational efficiency.\nIn this workshop we invite contributions from
  researchers working in the following thematic areas:\n\nMultilevel and Mu
 ltifidelity methods: Multilevel and multifidelity strategies judiciously 
 allocate computational resources between expensive high-fidelity models an
 d less costly low-fidelity models. These strategies allow for the efficien
 t computation of low variance/high accuracy results. Hence\, efficient sim
 ulation and UQ are possible.\nMultifidelity Surrogate Models: Resource co
 nstraints often limit data availability for computationally expensive simu
 lations. Multifidelity surrogate models\, such as multifidelity Gaussian p
 rocesses\, chain models with increasing fidelity\, balancing error and cos
 t. These surrogates support adaptive sampling strategies\, incrementally i
 mproving accuracy.\nApplications of Multifidelity Methods: Multifidelity t
 echniques are applied across disciplines\, from fusion energy to climate m
 odeling. This focus area will feature domain experts demonstrating the use
  of these tools in real-world scenarios\, showcasing their ability to prov
 ide accurate\, reliable UQ. This includes software packages for multifidel
 ity stochastic simulation and UQ.\n\n\nhttps://math-events.uni-bonn.de/eve
 nt/280/
LOCATION:Poppelsdorfer Allee 45\, 1. EG\, Lecture room (HIM)
URL:https://math-events.uni-bonn.de/event/280/
END:VEVENT
END:VCALENDAR