BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Scalable numerical linear algebra for science and machine learning
  [Lecture]
DTSTART:20261106T091500Z
DTEND:20261106T100000Z
DTSTAMP:20261030T013300Z
UID:indico-event-1712@math-events.uni-bonn.de
DESCRIPTION:Speakers: David Persson\n\nLarge-scale computational methods h
 ave become central to modern science. Methods from statistics\, machine l
 earning\, and physical simulation have transformed fields from weather pre
 diction to astrophysics to quantum mechanics. Despite the diversity of co
 mputational methods used in such fields\, a surprising number of our most
  important tools are based on a single shared computational core: numerica
 l linear algebra (NLA). Research in NLA impacts a huge range of downstream
  computational methods by improving algorithms for basic matrix computatio
 ns like linear system solving\, matrix multiplication\, and data compressi
 on. Among these\, matrix compression has become especially important for l
 arge-scale problems\, where storing or computing with a matrix directly m
 ay be prohibitively expensive. This talk focuses on replacing such matrice
 s by structured approximations that are cheaper to store and compute with
 . Randomization provides a powerful approach for computing these approxima
 tions\, even when a matrix is accessible only through matrix-vector produc
 ts. We will discuss recent results on hierarchical matrix compression\, 
 together with broader directions in NLA for scientific computing and machi
 ne learning. \n\nhttps://math-events.uni-bonn.de/event/1712/
LOCATION:Endenicher Allee 60/1-016 - Lipschitzsaal (Mathezentrum)
URL:https://math-events.uni-bonn.de/event/1712/
END:VEVENT
END:VCALENDAR