BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Nested sampling cross-checks using order statistics - Andew Fowlie
  (Nanjing Normal University)
DTSTART:20200715T100000Z
DTEND:20200715T103000Z
UID:TALK149575@talks.cam.ac.uk
CONTACT:Will Handley
DESCRIPTION:Nested sampling (NS) is an invaluable tool in data analysis in
  modern astrophysics\, cosmology\, gravitational wave astronomy and partic
 le physics. We identify a previously unused property of NS related to orde
 r statistics: the insertion indexes of new live points into the existing l
 ive points should be uniformly distributed. This observation enabled us to
  create a novel cross-check of single NS runs. The tests can detect when a
 n NS run failed to sample new live points from the constrained prior and p
 lateaus in the likelihood function\, which break an assumption of NS and t
 hus leads to unreliable results. We applied our cross-check to NS runs on 
 toy functions with known analytic results in 2 - 50 dimensions\, showing t
 hat our approach can detect problematic runs on a variety of likelihoods\,
  settings and dimensions. As an example of a realistic application\, we cr
 oss-checked NS runs performed in the context of cosmological model selecti
 on. Since the cross-check is simple\, we recommend that it become a mandat
 ory test for every applicable NS run.\n\nhttps://arxiv.org/abs/2006.03371
LOCATION:Battcock Tea area
END:VEVENT
END:VCALENDAR