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:Signal Processing and Communications Lab Seminars
SUMMARY:INDEPENDENT COMPONENT ANALYSIS VIA NONPARAMETRIC M
AXIMUM LIKELIHOOD ESTIMATION - Prof. Richard Samwo
rth\, Statistical Laboratory\, University of Cambr
idge
DTSTART;TZID=Europe/London:20131205T140000
DTEND;TZID=Europe/London:20131205T150000
UID:TALK47430AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/47430
DESCRIPTION:*Abstract*: Independent Component Analysis (ICA) m
odels are very popular semi-parametric models in w
hich we observe independent copies of a random vec
tor X = AS\, where A is a non-singular matrix and
S has independent components. We propose a new way
of estimating the unmixing matrix W = A^-1 and th
e marginal distributions of the components of S us
ing nonparametric maximum likelihood. Specifically
\, we study the projection of the emipirical distr
ibution onto the subset of ICA distributions havin
g log-concave marginals. We show that\, from the p
oint of view of estimating the unmixing matrix\, i
t makes no difference whether or not the log-conca
vity is correctly specified. The approach is furth
er justified by both theoretical results and a sim
ulation study.\n\n*Bio*: Richard Samworth obtain
ed his PhD in Statistics from the University of Ca
mbridge in 2004. Following a research fellowship a
t St John's College\, Cambridge\, he was appointed
to a lectureship in Statistics at the Statistical
Laboratory in Cambridge in 2005. He was promoted
to a readership in 2010 and to a full professorsh
ip from October 2013. Richard remains a fellow of
St John's College\, and currently holds an EPSRC E
arly Career Fellowship (worth GBP 1.2M) for five y
ears from December 2012.\n\nHis main research inte
rests are in nonparametric and high-dimensional st
atistics. Particular topics include shape-constrai
ned density and other nonparametric function estim
ation problems\, nonparametric classification\, cl
ustering and regression\, Independent Component An
alysis\, the bootstrap and high-dimensional variab
le selection problems. He was awarded the Royal S
tatistical Society Research prize (2008)\, a Lever
hulme Research Fellowship (2011) and the Royal Sta
tistical Society Guy Medal in Bronze (2012). He c
urrently serves as an Associate Editor for the Ann
als of Statistics\, the Journal of the Royal Stati
stical Society Series B\, Biometrika and Statistic
a Sinica.
LOCATION:LR5\, Engineering\, Department of
CONTACT:Prof. Ramji Venkataramanan
END:VEVENT
END:VCALENDAR