Agents in Biology
201
FIGURE 7.11: Using shortest possible error between the simulated results.
• Global values can influence the results produced. Global values can be
changed dynamically during the course of the model simulation. This
can essentially be done in both kinds of models depending on how the
models are written.
• Starting conditions of the model can have an effect on the model results.
This is seen in both approaches.
• Introducing dynamic inputs to the model. In an agent-based model,
dynamic agents can be introduced which get activated or influence the
progression of results. This can easily be programmed by having an
agent added which performs certain activities at certain time steps. This
would however be tedious to be programmed in an equation model as
complicated nested for-loops may need to be added to the model to allow
this. This involves very little changes in an agent-based model.
• Increasing complexity. Further complexity can be easily introduced in
agent-based models by adding agents and new functions. In equation
models, this would require rewriting of the equations and the source
code.
• Directed behavior. Agents are autonomous, goal-directed and sociable
elements. The decisions they make are based on bounded rationality
which means that each agent would have a sphere of influence which allows proximity to be checked before making decisions. In equation models this concept is not present. Here a list is traversed and everything in
201
FIGURE 7.11: Using shortest possible error between the simulated results.
• Global values can influence the results produced. Global values can be
changed dynamically during the course of the model simulation. This
can essentially be done in both kinds of models depending on how the
models are written.
• Starting conditions of the model can have an effect on the model results.
This is seen in both approaches.
• Introducing dynamic inputs to the model. In an agent-based model,
dynamic agents can be introduced which get activated or influence the
progression of results. This can easily be programmed by having an
agent added which performs certain activities at certain time steps. This
would however be tedious to be programmed in an equation model as
complicated nested for-loops may need to be added to the model to allow
this. This involves very little changes in an agent-based model.
• Increasing complexity. Further complexity can be easily introduced in
agent-based models by adding agents and new functions. In equation
models, this would require rewriting of the equations and the source
code.
• Directed behavior. Agents are autonomous, goal-directed and sociable
elements. The decisions they make are based on bounded rationality
which means that each agent would have a sphere of influence which allows proximity to be checked before making decisions. In equation models this concept is not present. Here a list is traversed and everything in