Abstract

Game theoretic models have proven to be powerful to gain insight into the processes of strategic decision-making and learning among interacting autonomous agents. I present a review of some fundamental concepts, emerging research, and open problems related to the analysis and control of evolutionary games, with particular emphasis on applications in social, economic, biological and robotic networks. In populations of autonomous agents, when individuals? self-interested goals conflict with the greater interest of the group, counter-intuitive outcomes and social dilemmas may arise. Evolutionary game theory has emerged as a vital tool set in the investigation of such network dynamics. In addition, one may explore how agents might learn to choose their strategies over time to adapt with the peers and surroundings; control theorists are particularly interested in knowing whether a small group of agents can manipulate the collective actions at large. Hence, decision- making, learning and control can be discussed in a unified framework, leading to new challenges and research opportunities for control scientists and engineers.