George K. Georgiou and Ya Ling Chang,
both Ph.D. candidates at the Graduate School and
University Center of the City University of New York, have been investigating the effectiveness
of Belief Networks to serve as the inference engines of such intelligent information systems. A
belief network is a mathematical tool that models the interdependencies of a set of events and is
used for computing the probability of events, given the probabilistic changes in other related events.
Such models are popular among the AI community, and have been used in diagnosis and prediction
systems. Mr. Georgiou is applying a pattern recognition technique to extract statistical relations
from temporal databases. These relations form the basis for automating the process of building such
networks.Currently Ms. Chang is exploring various techniques to simplify the structure of a belief
network while preserving its computational capabilities.
Click here to go to the Home Page of the Intelligent Computation Technology Club @ Queens College (CUNY)