BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:A New Machine Learning Library - Christian Steinruecken (Universit
 y of Cambridge)
DTSTART:20101213T111500Z
DTEND:20101213T120000Z
UID:TALK28444@talks.cam.ac.uk
CONTACT:Emli-Mari Nel
DESCRIPTION:This will be an informal talk about a new machine learning lib
 rary written (from scratch) in Java.  I will talk about some of its nice p
 roperties\, e.g. how it makes use of Java generics.  I will show some nice
  example code\, e.g. [1] fully generic sampling algorithms (Metropolis\, e
 tc.) which can be instantiated in 1 line of code for any space and distrib
 ution\; [2] several elegant ways of constructing fully generic Pitman-Yor 
 and Dirichlet-Processes\, such that samples can themselves be used as dist
 ributions\; [3] interesting dualities between stick-breaking constructions
  and distributions over natural numbers\;  [4] how to define distributions
  over arbitrary grammars in very few lines of code.
LOCATION:TCM Seminar Room\, Cavendish Laboratory\, Department of Physics
END:VEVENT
END:VCALENDAR