Soldier Behavior

We apply agent-based models to examine soldiers conducting tactical operations to evaluate plans of engagement under different conditions. An agent is a self-contained software program that performs a task by interacting with other agents and with its environment. A smart agent is an agent that uses learning algorithms to determine how to optimally perform the task that it is given. Our premise is that advanced learning models like genetic programming (GP) can be used to develop smart agents that mimic the complex decision process of humans. Our goal is to develop a GP framework for building behavioral models of soldiers that can be used in war gaming simulations. Contact: Mark Boslough, 505-845-8851; mbboslo@sandia.gov