BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Applications of Algorithmic Information Theory - Marcus Hutter (De
 epMind)
DTSTART:20260218T140000Z
DTEND:20260218T150000Z
UID:TALK244663@talks.cam.ac.uk
CONTACT:Georg Maierhofer
DESCRIPTION:Algorithmic information theory has a wide range of application
 s\, despite the fact that its core quantity\, Kolmogorov complexity\, is i
 ncomputable. Most importantly\, AIT allows to quantify Occam’s razor\, t
 he core scientific paradigm that ”among two models that describe the dat
 a equally well\, the simpler one should be preferred”. This led to unive
 rsal theories of induction and action in the field of machine learning and
  artificial intelligence\, and practical versions like the Minimum Encodin
 g Length (MDL/MML) principles. The universal similarity metric probably sp
 awned the greatest practical success of AIT. Approximated by standard comp
 ressors like Lempel-Ziv (zip) or bzip2 or PPMZ\, it leads to the normalize
 d compression distance\, which has been used to fully automatically recons
 truct language and phylogenetic trees\, and many other clustering problems
 . AIT has been applied in disciplines as remote as Cognitive Sciences\, Bi
 ology\, Physics\, and Economics.
LOCATION:Centre for Mathematical Sciences\, MR5
END:VEVENT
END:VCALENDAR