BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Non-parametric Bayesian Method and Maximum-A-Posteriori Inference 
 in Statistical Machine Translation - Tsuyoshi Okita (Dublin City Universit
 y)
DTSTART:20120502T101500Z
DTEND:20120502T111500Z
UID:TALK38018@talks.cam.ac.uk
CONTACT:Zoubin Ghahramani
DESCRIPTION:Since recent sophisticated Machine Learning algorithms implici
 tly handle various things\, practitioners do not need to worry much about 
 how to deploy those algorithms in particular situations. However\, if it c
 omes to real-life data such as Statistical Machine Translation\, several t
 hings were worth considering: 1) the underlying distribution may be better
  assumed to be the power-law distribution rather than its i.i.d. counterpa
 rt\, 2) noise may not be captured well as a simple Gaussian type (hence\, 
 such noise assumption is not often embedded in the ML algorithm)\, 3) avai
 lable prior knowledge may not be sufficiently used\, and so forth. It is n
 oted that what kinds of non-Gaussian type noise we need to focus on and wh
 at kind of prior knowledge we need to target were not evident from the beg
 inning (These issues would be quite difficult even if we can exploit the d
 omain experts. This is since these require both the knowledge of the under
 lying ML algorithm and the domain knowledge of the area). We discuss two a
 lgorithms in the application area of Statistical Machine Translation: non-
 parametric Bayesian method (hierarchical Pitman-Yor process related topics
 ) and Maximum-A-Posteriori inference. The first algorithm is related to th
 e language model smoothing where 1) is concerned\, while the second algori
 thm is related to the word alignment where 2) and 3) are concerned.
LOCATION:Engineering Department\, CBL Room BE-438
END:VEVENT
END:VCALENDAR