A Brief Description
of Our Queueing Network Modeling Work:
We model the human mind and behavioral system
as a queueing network of information processing servers. Our modeling approach
integrates theories and methods of mathematical psychology, computational
cognitive modeling (procedure/production rule based), operations research,
neuroscience, and experimental psychology.
As a mathematical
model, our queueing network model is able to integrate a large number of
existing mathematical models as special cases and to cover a wider range of
phenomena. As a computational cognitive model, our model has a distinct
position in modeling cognition: Cognition in our Queueing Network model is
neither serial (ACT-R's position) nor relying on Executive Control (EPIC's
position). It is a macro-architecture model with a few dozens of powerful
processing units (like cortical zones or psychological functional modules), not
a micro-structure connectionist neural network of a large number of neurons.
Our queueing network model has not only an underlying mathematical structure
and rigor, but also computational power to generate/simulate behavior in real
time. For example, it is able to steer a driving simulator in real time, while
allowing us to visualize the internal information flows inside "the
mind" while the "human" performs real world tasks such as
driving--see short
movie clip (Best viewed with QuickTime).
Please see our publications on Queueing Network Modeling of Human Performance for more details. We warmly welcome your comments.
Journal Articles: (Please click the Journal Publications Link on my Homepage)