BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Sabre//Sabre VObject 4.6.1//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:Europe/Zurich
X-LIC-LOCATION:Europe/Zurich
TZURL:http://tzurl.org/zoneinfo/Europe/Zurich
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19810329T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19961027T030000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:news2124@dmi.unibas.ch
DTSTAMP;TZID=Europe/Zurich:20261115T172224
DTSTART;TZID=Europe/Zurich:20261120T110000
SUMMARY:Seminar in Numerical Analysis: Daniel Kressner (EPFL)
DESCRIPTION:Randomized algorithms are becoming increasingly popular in matr
 ix computations. Recent software efforts\, such as RandLAPACK\, demonstrat
 e that randomization is on the verge of replacing existing deterministic t
 echniques for several large-scale linear algebra tasks in scientific compu
 ting. The poster child of these developments\, randomized SVD\, is now one
  of the state-of-the-art approaches for performing low-rank approximation.
  In this talk\, we will go beyond the randomized SVD and illustrate the gr
 eat potential of randomization to not only speed up existing algorithms\, 
 but to also yield novel and often simple algorithms for solving notoriousl
 y difficult problems. Examples covered in this talk include reduced order 
 modeling\, acceleration of Krylov subspace methods\, joint diagonalization
 \, large null space computation\, spectral density estimation\, and tensor
  structured randomness. A common theme of these developments is that rando
 mization helps transform linear algebra results that only hold generically
  into robust and reliable numerical algorithms. To a large extent\, this t
 alk reflects the themes covered in the recent Acta Numerica survey https:/
 /arxiv.org/abs/2610.38157 \, co-authored with Ethan Epperly and Yuji Nakat
 sukasa.\\r\\nFor further information about the seminar\, please visit this
  webpage [https://dmi.unibas.ch/de/forschung/mathematik/seminar-in-numeric
 al-analysis/].
X-ALT-DESC:<p>Randomized algorithms are becoming increasingly popular in ma
 trix computations. Recent software efforts\, such as RandLAPACK\, demonstr
 ate that randomization is on the verge of replacing existing deterministic
  techniques for several large-scale linear algebra tasks in scientific com
 puting. The poster child of these developments\, randomized SVD\, is now o
 ne of the state-of-the-art approaches for performing low-rank approximatio
 n. In this talk\, we will go beyond the randomized SVD and illustrate the 
 great potential of randomization to not only speed up existing algorithms\
 , but to also yield novel and often simple algorithms for solving notoriou
 sly difficult problems. Examples covered in this talk include reduced orde
 r modeling\, acceleration of Krylov subspace methods\, joint diagonalizati
 on\, large null space computation\, spectral density estimation\, and tens
 or structured randomness. A common theme of these developments is that ran
 domization helps transform linear algebra results that only hold generical
 ly into robust and reliable numerical algorithms. To a large extent\, this
  talk reflects the themes covered in the recent Acta Numerica survey https
 ://arxiv.org/abs/2610.38157 \, co-authored with Ethan Epperly and Yuji Nak
 atsukasa.</p>\n<p>For further information about the seminar\, please visit
  this <a href="https://dmi.unibas.ch/de/forschung/mathematik/seminar-in-nu
 merical-analysis/">webpage</a>.</p>
DTEND;TZID=Europe/Zurich:20261120T123000
END:VEVENT
END:VCALENDAR
