BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Joydeep Chowdhury "A test for multi-sample comparison for infinite
dimensional data based on spatial signs"
DTSTART;VALUE=DATE-TIME:20191204T093000Z
DTEND;VALUE=DATE-TIME:20191204T103000Z
DTSTAMP;VALUE=DATE-TIME:20191209T102755Z
UID:indico-event-5358@indico.math.cnrs.fr
DESCRIPTION:We consider an ANOVA type multi-sample comparison problem\, wh
ere the observations are infinite dimensional in nature. In particular\, t
he observations may be random functions. We propose a test based on spatia
l signs in such a setup. An asymptotic implementation and a bootstrap impl
ementation of this test are developed\, and the asymptotic consistency of
both procedures is established. We compare the finite-sample and asymptoti
c performances of our test with that of several other tests of ANOVA for f
unctional data in the literature. We found that our test not only outperfo
rms those tests in several non-Gaussian models with heavy tails\, but in s
ome Gaussian models also\, it exhibits better performance than them.\n\nht
tps://indico.math.cnrs.fr/event/5358/
LOCATION:IMB 318
URL:https://indico.math.cnrs.fr/event/5358/
END:VEVENT
END:VCALENDAR