Agent-based Models are computer programs that try to simulate various complex phenomena via virtual "agents" that represent the components of a business system. The behaviors of these agents are programmed with simple rules. When put into motion, the models realistically depict how business is conducted through "emergence". Programming approaches can range from the use of "if-then" rules as in expert systems to sophisticated machine learning algorithms such as neural networks that allow agents to learn and modify their behavior during the simulation. Parameters of the model are set to represent a situation of interest, and the model is run for thousands of iterations. Agent-Based Models are calibrated against historical data to ensure the model is accurately replicating the behavior of the real system. Agent-based Models can provide valuable information about the dynamics of the real-world system that they emulate. As widely varied individual agents interact in the model, the simulation shows how their collective behaviors govern the performance of the entire system - for instance, the emergence of a successful product or an optimal schedule. These simulations are powerful strategic tools for "what-if" scenario analysis: as managers change agent characteristics or "rules," the impact of the change can be easily seen in the model output. Perhaps most important, the computer can generate strategies that the user might never have considered.

Agent-Based Modeling Whitepaper