BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Isaac Newton Institute Seminar Series
SUMMARY:Functional regression on manifold with contaminati
on - Fang Yao (University of Toronto\; University
of Toronto)
DTSTART;TZID=Europe/London:20180320T160000
DTEND;TZID=Europe/London:20180320T170000
UID:TALK102682AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/102682
DESCRIPTION:We propose a new perspective on functional regress
ion with a predictor process via the concept of ma
nifold that is intrinsically finite-dimensional an
d embedded in an infinite-dimensional functional s
pace\, where the predictor is contaminated with di
screte/noisy measurements. By a novel method of
functional local linear manifold smoothing\, we ac
hieve a polynomial rate of convergence that adapts
to the intrinsic manifold dimension and the level
of sampling/noise contamination with a phase tran
sition phenomenon depending on their interplay. Th
is is in contrast to the logarithmic convergence r
ate in the literature of functional nonparametric
regression. We demonstrate that the proposed metho
d enjoys favorable finite sample performance relat
ive to commonly used methods via simulated and rea
l data examples. (Joint with Zhenhua Lin)
LOCATION:Seminar Room 1\, Newton Institute
CONTACT:INI IT
END:VEVENT
END:VCALENDAR