My work with multi-agent systems goes back to 2015! I earned my Master's degree in Computer Science specializing in these systems at the University of Southern California.

I was supported by a fellowship from the Department of Homeland Security. I became somewhat (understandably) jaded about the nature of the work in 2016.* 

My contribution to the literature was to consider a special case of game theory formulation known as a Stackelberg game. In this game, there are two players who are competing for utility and each act and subsequently react to one another, in turn.

The two players that are participating are known as agents-- one agent represents the attacker, or any party that poses a threat, or seeks to evade detection by the other, defender agent. This specific game was posed as zero-sum, and was applied to the Transportation Security Administration (TSA). Previous work had not considered the uncertain nature of airport screening resource allocation-- we don't know when the passengers will arrive, and we don't have unlimited access to the most effective screening resources either. My research focused on designing various algorithms to ensure strong defender performance in spite of their state of imperfect information. 

Though I left the project, the work continued and was published shortly after my departure. You can read the paper here.

*The program was widey advertised as "AI for Social Good" by my advisor, and my work alone would have made an excellent chapter for Cathy O'Neill's "Weapons of Math Destruction." I spent the next few years after my time in graduate school focusing on education.