BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Lifted Message Passing: A Step Towards Gaining a 'Big Picture' Vie
 w on AI - Kristian Kersting
DTSTART:20101130T100000Z
DTEND:20101130T110000Z
UID:TALK28118@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:Many AI inference problems arising in a wide variety of fields
  such as machine learning\, semantic web\, network communication\, compute
 r vision\, and robotics can be solved using message-passing algorithms tha
 t operate on factor graphs. Often\, however\, we are facing inference prob
 lems with symmetries not reflected in the factor graph structure and\, hen
 ce\, not exploitable by efficient message-passing algorithms. For instance
 \, unifying first-order logic and probability is a long-standing goal of A
 I\, and in recent years many representations combining aspects of the two 
 have been proposed. In the inference stage\, however\, they still operate 
 on a mostly propositional representation level and do not exploit the addi
 tionaly symmetries  often induced. \n\nIn this talk\, I shall present lift
 ed message-passing algorithms that exploit such additional symmetries. Sta
 rting from a given factor graph\, they essentially first construct a lifte
 d factor graph of supernodes and superfactors\, corresponding to sets of n
 odes and factors that send and receive the same messages\, i.e.\, that are
  indistinguishable given the evidence. Then\, they run a modified message-
 passing algorithm on the lifted factor. In particular\, I shall present li
 fted variants of loopy and Gaussian belief propagation as well as warning 
 and survey propagation\, and demonstrate that significant efficiency gains
  are obtainable\, often by orders of magnitude. These contributions advanc
 e the theoretical understanding of inference within large probabilistic mo
 dels. More importantly\, they put a 'Big Picture' view on AI in reach that
  can be called "Statistical Relational AI".\n
LOCATION:Small lecture theatre\, Microsoft Research Ltd\, 7 J J Thomson Av
 enue (Off Madingley Road)\, Cambridge
END:VEVENT
END:VCALENDAR