22
X-Machines for Agent-Based Modeling: FLAME Perspectives
communication mechanisms to form patterns. Stigmergy is a communication
mechanism insects use to interact with each other using the environment.
For example, ants use pheromones to communicate pathways with other ants.
Human societies use messages to communicate information to each other.
Similarly, multi-agent systems use communication for coordination in a selforganizing system.
2.2.1 Bring in the Agents
Modeling of complex system behavior is an emergent science which demonstrate complex or social behavior of different communities. Agent-based modeling is a technique which best models these systems, an alternative to conventional differential equation methods. This approach allows a bottom-up
procedure, where the focus concentrates on individual interacting units, given
clear defined rules and allowed to simulate. The produced emergent pattern
of system behavior can then be studied to test and understand behavior of
complex systems, otherwise not possible from studying from an outside view.
There are various agent-based environments that can be used to design and
test models. Each of these are based on different computational models, varying in computational languages used. Grimm et al. [77] discussed a detailed
overview of the problems of verifying models because the tools themselves,
are not being designed on predefined software methodologies. They recognize
a need for rules to creating agent-based models. Generalizing these rules, allows models to be created with formal methods, encouraging credibility of the
results.
Figure 1.5 discussed the process involved in writing an agent-based model.
The model starts with a description about individual elements as agents. These
agents are using a set of memory variables, functions and communication protocols, that allow them to communicate with each other and the environment.
Agents are implemented as separate pieces of code, which communicate using
messages.
The individual agent interactions allow certain macro variables to emerge
in the system, depicting how whole systems collectively behave. The simulated
model can be tested against real data to check its accuracy and validation.
However, the complexity in agent-based models increases as,
• Agents can travel in space unlike agents or cells represented in layers of
cellular automata.
• Every agent may have limitations in cognitive, physical or temporal
abilities based on the model.
• The interaction dynamics between agents lead to emergent patterns to
mimic natural system behaviors [174].
• Adopting agent-oriented approaches to natural systems involves model-
X-Machines for Agent-Based Modeling: FLAME Perspectives
communication mechanisms to form patterns. Stigmergy is a communication
mechanism insects use to interact with each other using the environment.
For example, ants use pheromones to communicate pathways with other ants.
Human societies use messages to communicate information to each other.
Similarly, multi-agent systems use communication for coordination in a selforganizing system.
2.2.1 Bring in the Agents
Modeling of complex system behavior is an emergent science which demonstrate complex or social behavior of different communities. Agent-based modeling is a technique which best models these systems, an alternative to conventional differential equation methods. This approach allows a bottom-up
procedure, where the focus concentrates on individual interacting units, given
clear defined rules and allowed to simulate. The produced emergent pattern
of system behavior can then be studied to test and understand behavior of
complex systems, otherwise not possible from studying from an outside view.
There are various agent-based environments that can be used to design and
test models. Each of these are based on different computational models, varying in computational languages used. Grimm et al. [77] discussed a detailed
overview of the problems of verifying models because the tools themselves,
are not being designed on predefined software methodologies. They recognize
a need for rules to creating agent-based models. Generalizing these rules, allows models to be created with formal methods, encouraging credibility of the
results.
Figure 1.5 discussed the process involved in writing an agent-based model.
The model starts with a description about individual elements as agents. These
agents are using a set of memory variables, functions and communication protocols, that allow them to communicate with each other and the environment.
Agents are implemented as separate pieces of code, which communicate using
messages.
The individual agent interactions allow certain macro variables to emerge
in the system, depicting how whole systems collectively behave. The simulated
model can be tested against real data to check its accuracy and validation.
However, the complexity in agent-based models increases as,
• Agents can travel in space unlike agents or cells represented in layers of
cellular automata.
• Every agent may have limitations in cognitive, physical or temporal
abilities based on the model.
• The interaction dynamics between agents lead to emergent patterns to
mimic natural system behaviors [174].
• Adopting agent-oriented approaches to natural systems involves model-