BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Evidence is all you need: Nested Sampling for particle physics - D
 avid Yallup (Cambridge)
DTSTART:20221206T110000Z
DTEND:20221206T120000Z
UID:TALK183620@talks.cam.ac.uk
CONTACT:William Fawcett
DESCRIPTION:The Bayesian vs. Frequentist debate has raged for many years i
 n particle physics\, with the battle lines largely drawn and a set of esta
 blished techniques falling firmly in favour of frequentist thinking. The s
 ame is largely true on the cosmology side of the fence\, where the dominan
 t paradigm is decidedly Bayesian. This is a curious fact that is worth con
 sidering\, how have two adjacent fields that are so similar in ultimate go
 als and theoretical basis ended up at loggerheads when it comes to inferen
 ce? To try and unravel some of this mystery\, this talk will discuss the '
 Rosetta Stone' of inference tasks in particle physics - searching for bump
 s in smoothly falling spectra at colliders. I will aim to extract the sign
 al (the higgs) from the noise (the diphoton background)\, however in doing
  so I will really be attempting to extract the signal (why we use Nested S
 ampling so much in Cosmology) from the noise (the Bayesian vs. Frequentist
  debate). An "evidence based" approach (powered by Nested Sampling) afford
 s unique perspectives on some long standing inference problems for High En
 ergy Physics.
LOCATION:Ryle Seminar Room
END:VEVENT
END:VCALENDAR