|
|
|
|
What are agent-based models?
Agent-based models consist of two interacting
components, an environment and a collection of agents. When
used in the context of land-use/land-cover change, the
landscape is divided into discrete units (grid cells or
parcels), each with a particular land-use/cover type and
perhaps other attributes. A geographic information system
is often used to implement the landscape portion of the
model. Agents are independent entities with individual
characteristics and histories. Each agent may interact with
other agents through communication mechanisms and with the
environment through a decision-making process that both
influences and is influenced by the environment (Parker
et al. 2003; Bosquet and
Le Page 2004). The models incorporate heterogeneity
both across the landscape and among agents (Parker
et al. 2003). This heterogeneity and the feedbacks
incorporated in the interactions between the components of
the model are the basis of a complex system. While the
individual landscape elements and agents are relatively
simple entities, their interdependent behavior leads to
emergent properties of the system that can be studied using
an agent-based model (Bosquet
and Le Page 2004).
|
|
|
The articles included in this annotated
bibliography are divided into three categories:
-
Overviews: Articles that provide general
descriptions of agent-based models and review their
applications
-
Example
Models: Articles that describe a specific
agent-based model and how it has been used
-
Other
Issues Relevant to Agent-Based Modeling:
Articles that deal with issues such as defining agents,
integrating ABMs and GIS, and abstract experiments
conducted with ABMs.
Other Web Sites of Interest
Ecology and Society Special Issue: The online journal
Ecology and Society published a special issue on
"Empirically Based, Agent-based Modeling of
Social-Ecological Issues" in December 2006.
Project SLUCE:
Project SLUCE (Spatial Land
Use Change and Ecological Effects at the
Rural-Urban Interface) is a group based at the University of
Michigan that uses agent-based modeling to investigate
land-use. The principle investigator is Dan Brown,
co-author of many of the articles discussed on the
issues page of this bibliography.
The project's web page includes both publications and code
and documentation for many of the models created by the
group.
|
