Agent Based Modeling (ABM)

• This project involves refining and extending agent based modeling to simulate the ontology and morphology of young children’s playgroups.
  
  

• ABM is a model that has agents (e.g., children, ants) implementing local interactions according to a set of algorithms (i.e., rules) constructed by the investigator consistent with theory or observed data. Simulations are tools to enrich scientific understanding of fundamental processes.
  
  

• Agents are autonomous and interact with other agents and their environment. Their actions are goal directed and they evolve upon multiple iterations. These characteristics permit the agents to engage in complex interactions unguided by the investigator.
  
  

• Abduction, inference to the best explanation, is used to infer the generating mechanisms of real world processes; in turn, these inferences are used to articulate the rules of an ABM.
  
  

• Each agent possesses a strategy set that is invoked during interactions with other agents; agent-to-agent interaction determines and is reciprocally influenced by the emergence of structure, as agent and structure co-evolve within the environment. Repetitive interactions among heterogeneous agents are studied to examine the emerging structures that evolve for the particular set of rules implemented by the agents.
  

• As exchanges and interactions occur in the simulation, agents and the rules they carry evolve, producing complex biological and social processes analogous to those found in the real world.
  
  

• An ABM generates output that represents the modeler’s best attempt to replicate reality.