PhD Seminar - Hongxuan Jin

Date: Thursday November 19, 2009

The School of Computer Science is pleased to announce:

 Heuristic Assignment of CPDs for Probabilistic Inference in Junction Trees

by Hongxuan Jin

Place:  CS Seminar Room, 3105 LT

Date:   Monday, November 23, 2009

Time:  10:00 - 11:00 AM

Abstract:

Extensive research has been done for efficient computation of probabilistic queries posed to Bayesian networks (BNs).  One popular architecture for exact inference on BNs is the Junction Tree (JT) based architecture.  Among all variations developed, HUGIN is the most efficient JT-based architecture.  The Global Propagation (GP) method used in the HUGIN architecture is arguably one of the best methods for probabilistic inference in BNs before the propagation, initialization is done to obtain the potential for each cluster in the JT.  Then with the GP method, each cluster potential is transformed into cluster marginal through passing messages with its neighbouring clusters.  Improvements have been proposed to make the message propagation more efficient.  Still, the GP method can be very slow for dense networks.  As BNs are applied to larger, more complex and realistic applications.  The design of more efficient inference algorithm has become increasingly important.  Towards this goal, we present a heuristic for initialization that avoids unnecessary message passing among clusters of a JT, therefore improving the performance of the architecture by passing less messages.