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:NLIP Seminar Series
SUMMARY:The Geometry of Machine Translation - Rory Waite\,
University of Cambridge
DTSTART;TZID=Europe/London:20150417T120000
DTEND;TZID=Europe/London:20150417T130000
UID:TALK58527AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/58527
DESCRIPTION:Most modern statistical machine translation system
s are based on linear statistical models. One extr
emely effective method for estimating the model pa
rameters is minimum error rate training (MERT)\, w
hich is an efficient form of line search adapted t
o the highly non-linear objective functions used i
n machine translation. We will show that MERT can
be represented using convex geometry\, which is th
e mathematics of polytopes and their faces. Using
this geometric representation of MERT we investiga
te whether the optimisation of linear models is tr
actable in general. It has been believed that the
number of feasible solutions of a linear model is
exponential with respect to the number of sentence
s used for parameter estimation\, however we show
that the exponential complexity is instead due to
the feature dimension. This result has important r
amifications because it suggests that the current
trend in building statistical machine translation
systems by introducing very large number of sparse
features is inherently not robust.
LOCATION:FW26\, Computer Laboratory
CONTACT:Tamara Polajnar
END:VEVENT
END:VCALENDAR