BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:The Geometry of Machine Translation - Rory Waite\, University of C
 ambridge
DTSTART:20150417T110000Z
DTEND:20150417T120000Z
UID:TALK58527@talks.cam.ac.uk
CONTACT:Tamara Polajnar
DESCRIPTION:Most modern statistical machine translation systems are based 
 on linear statistical models. One extremely effective method for estimatin
 g the model parameters is minimum error rate training (MERT)\, which is an
  efficient form of line search adapted to the highly non-linear objective 
 functions used in machine translation. We will show that MERT can be repre
 sented using convex geometry\, which is the mathematics of polytopes and t
 heir faces. Using this geometric representation of MERT we investigate whe
 ther the optimisation of linear models is tractable in general. It has bee
 n believed that the number of feasible solutions of a linear model is expo
 nential with respect to the number of sentences used for parameter estimat
 ion\, however we show that the exponential complexity is instead due to th
 e feature dimension. This result has important ramifications because it su
 ggests 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
END:VEVENT
END:VCALENDAR