What is Agent-Based Modeling?

Agent-based modeling, or ABM, is type of computational simulation where a system is modeled as a collection of autonomous decision-making entities called agents. Each agent individually evaluates its situation and environment to decide what to do next. The individual behaviors and actions by agents impact the overall system. The result is that complex and interesting behaviors of the entire system emerge over time. Agents can be any decision-making entity in the environment, from individual people, to households, to ships at sea. Interactions between agents are a common feature of ABMs, and the decisions made by agents are typically governed by a set of rules for making decisions or applying new conditions to the agents. Even the most simple models can evolve complex behavior patterns and provide useful insight into the real-world system that it emulates

A common approach for ABM is to have agents that represent individual people. Each agent has a set of demographics, social contacts, and interactions with an environment representing a real or imagined geographic location. This enables the model to include individual responses and behaviors.

Specific responses can vary according to the individual’s characteristics, as well as the interactions between members of various social groups, such as the individual's neighborhood, school, or workplace. Because individuals in such ABMs are located within a specific geographical space, the models can be used to investigate interactions between individuals and spatially distributed resources such as workplaces or health care facilities.

In summary, agent-based models can reveal how interactions among individuals and their environment can result in patterns of population behavior. This approach has been shown to be particularly useful in understanding or predicting the impact of public policy and programs on population health and population dynamics.