Where Industry Meets Innovation

  • Contact Us
  • sign in Sign In
  • Sign in with certificate
mit campus


Include past months.

  • View All
  • ILP Conferences
  • All Conferences
  • Sloan Executive Education
  • MIT Professional Education
  • All Courses
  • Lectures
  • Recruitment
  • Event Series:

MIT/ILP Calendar Event

Obtaining Semantics in Bayesian Network Inference

November 9, 2011, 10-11 AM


Cory J. Butz
University of Regina

Host: Regina Barzilay
Host Affiliation: MIT CSAIL

Variable Elimination (VE), proposed by Zhang and Poole, is a standard
algorithm for performing exact inference in discrete Bayesian networks. VE
starts and ends with clear semantics, yet the intermediate factors constructed
by VE are seen as not having semantics. In this seminar, we give an algorithm,
called Semantics in Inference (SI), that denotes the semantics of every
intermediate factor constructed by VE. We show that SI is correct in nearly
every instance of Bayesian network inference. SI provides a better
understanding of the theoretical foundation of Bayesian networks and can be
used for improved clarity, as shown via an examination of Bayesian network
literature. This work is also interesting in that it reveals that d-separation
plays a much larger role in inference than the literature suggests.

Cory J. Butz received his Ph.D. degree in computer science from the
University of Regina, Regina, SK, Canada, in 2000. He then joined the School
of Information Technology and Engineering at the University of Ottawa, Ottawa,
ON, Canada, as an Assistant Professor. In 2001, he returned to the Department
of Computer Science at the University of Regina, where he currently holds the
rank of Professor. His research findings on Bayesian networks have drawn
invitations to visit Google Inc., USA and the University of Cambridge, UK.


Building 32 Map

Building 32, Room G451


For more information please contact: Marcia Davidson, 617-253-3049, marcia@csail.mit.edu