Modeling individuals: Agent-based approaches

Many mathematical models assume that everyone reacts in the same way to a disease, ignoring differences in personality, risk assessment, or personal choice. This assumption exists in stark contract to many of the concerns in public health planning: will doctors, nurses and other medical professionals carry out their professional responsibilities during an epidemic, or will they stay home? How can we encourage individuals to isolate themselves, rather than going to work, to present the spread of flu? Why do some parents refuse to vaccinate their children, and what impact does this have on childhood diseases?

Agent-based models try to address these questions by modeling the population as a collection of individuals (“agents”), each with their own behaviors, biology and movement patterns. The use of these models to examine questions of behavior using the zombie outbreak example will be discussed, as will applications to real-world questions like outbreak interventions or school-closure strategies.

Learning Objectives:

1. Describe an agent-based model.
2. Explain the differences between an agent-based model and a network or compartmental model.
3. Identify agent parameters of particular relevance to the learner’s research interests.