Speaker
Prof.
Misha Chertkov
(LANL)
Description
Loop Calculus introduced in [Chertkov, Chernyak '06]
constitutes a new theoretical tool that expresses explicitly
the symbol Maximum-A-Posteriori solution of a general
statistical inference problem via a solution of the Belief
Propagation, or Bethe-Pieirls, equations. This
finding brought a new significance to the BP concept, which
in the past was thought of as just a loop-free
approximation. In this presentation I will explain main
concept, algorithmic implications and feature challenges of
the Loop Calculus approach.
