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:Statistics
SUMMARY:Testing lack-of-fit in inverse regression models -
Gerda Claeskens (K.U.Leuven)
DTSTART;TZID=Europe/London:20080530T140000
DTEND;TZID=Europe/London:20080530T150000
UID:TALK11790AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/11790
DESCRIPTION:We propose two test statistics for use in inverse
regression problems where only noisy\, indirect ob
servations for the mean function are available. Bo
th test statistics have a counterpart in classical
hypothesis testing\, where they are called the or
der selection test and the data-driven Neyman smoo
th test. We also introduce two model selection cri
teria which extend the classical AIC and BIC to in
verse regression problems. In a simulation study w
e show that the inverse order selection and Neyman
smooth tests outperform their direct counterparts
in many cases. The methods are applied to data ar
ising in confocal fluorescence microscopy. Here\,
images are observed with blurring (modeled as deco
nvolution) and stochastic error at subsequent time
s. The aim is then to reduce the signal to noise r
atio by averaging over the distinct images. In thi
s context it is relevant to test whether the image
s are still equal (or have changed by outside infl
uences such as moving of the object table). This i
s joint work with N. Bissantz\, H. Holzmann and A.
Munk. \n
LOCATION:MR12\, CMS\, Wilberforce Road\, Cambridge\, CB3 0W
B
CONTACT:
END:VEVENT
END:VCALENDAR