Abstract
Validation of simulation models is extremely important.
It ensures that the right model has been built and lends
confidence to the use of that model to inform critical
decisions. Agent-based models (ABM) have been widely
deployed in different fields for studying the collective
behavior of large numbers of interacting agents. However,
researchers have only recently started to consider the issues
of validation. Compared to other simulation models, ABM
has many differences in model development, usage and
validation. An ABM is inherently easier to build than a
classical simulation, but more difficult to describe formally
since they are closer to human cognition. Using multi-agent
models to study complex systems has attracted criticisms
because of the challenges involved in their validation [1]. In
this report, we describe the challenge of ABM validation and
present a novel approach we recently developed for an ABM
system.