University of Houston
Department of Computer Science

In partial fulfillment of the Requirements for the Degree of
Master of Science

Chinfu Chou
will defend his thesis

DESIGN AND IMPLEMENTATION OF A
MULTIPLE BELIEF NETWORK TOOL SUITE

Abstract

Many efforts in artificial-intelligence research have been focused on modeling and reasoning with uncertainty in knowledge-based systems.  Bayesian belief network is one of the methods that have been successfully used for automating decision making in complex domains. Current belief network development environments support the design and execution of a single belief network; however, for certain applications it is desirable to be able to execute multiple belief networks in parallel. Applications that need this capability include sensitivity analysis, debugging, multi-expert decision making, and implementing generalized belief networks. In this thesis, a multiple belief execution and analysis tool suite is designed and implemented. This tool suite can be used to analyze the behavior of multiple belief networks and produces summaries using color html-displays. Special measures to analyze similarities and differences in belief network behavior have been provided. The tool suite also supports the comparison of a belief network ensemble with a reference belief network. Moreover, results of case studies are presented that used the tool suite for analyzing belief networks for diagnosing Huntington disease.

Date: Wednesday, October 3, 2001
Time: 2:00 PM
Place: 550-PGH

Faculty, students, and the general public are invited.
Thesis Advisor: Dr. Christoph F. Eick