Title :
RedTNet: A network model for strategy games
Author :
Hingston, Philip ; Preuss, Mike ; Spierling, Daniel
Author_Institution :
Sch. of Comput. Sci. & Security, Edith Cowan Univ., Perth, WA, Australia
Abstract :
In this work, we develop a simple, graph-based framework, RedTNet, for computational modeling of strategy games and simulations. The framework applies the concept of red teaming as a means by which to explore alternative strategies. We show how the model supports computer-based red teaming in several applications: realtime strategy games and critical infrastructure protection, using an evolutionary algorithm to automatically detect good and often surprising strategies.
Keywords :
computer games; evolutionary computation; games of skill; graph theory; real-time systems; RedTNet; computational modeling; computer-based red teaming; critical infrastructure protection; evolutionary algorithm; graph-based framework; realtime strategy games; Artificial intelligence; Computational modeling; Context; Evolutionary computation; Games; Optimization; Subspace constraints;
Conference_Titel :
Evolutionary Computation (CEC), 2010 IEEE Congress on
Conference_Location :
Barcelona
Print_ISBN :
978-1-4244-6909-3
DOI :
10.1109/CEC.2010.5586505