Biological swarms utilize simple behaviors spread over many agents to protect and promote the health of the colony. Recent interest in distributing tasks across swarms of autonomous agents has exposed the complexity of communications, task assignment, distributed sensing, and mission replanning for large teams of robotic agents. In this paper, we describe the approach we have taken to characterize and model the individual and collaborative behaviors used by Temnothorax rugatulus ants amid competitive nest selection scenarios. This species of ants exhibit favorable characteristics self-organizing without need for centralized planning, control, or communication. This work focuses on learning behavior rules from these ants to help design responses to swarm-based combat operations. This research effort consisted of three main parts: data generation, behavior characterization, and machine learning for UAV applications.