This paper describes the EvoTanks research project, a continuing attempt to develop strong AI players for a primitive 'Combat' style video game using evolutionary computational methods with artificial neural networks. A small but challenging feat due to the necessity for agent's actions to rely heavily on opponent behaviour. Previous investigation has shown the agents are capable of developing high performance behaviours by evolving against scripted opponents; however these are local to the trained opponent. The focus of this paper shows results from the use of co-evolution on the same population. Results show agents no longer succumb to trappings of local maxima within the search space and are capable of converging on high fitness behaviours local to their population without the use of scripted opponents.
|Number of pages||6|
|Publication status||Published - 2007|
|Event||IEEE 2007 Symposium on Computational Intelligence and Games (CIG '07) - Hawaii, USA|
Duration: 1 Apr 2007 → 5 Apr 2007
|Conference||IEEE 2007 Symposium on Computational Intelligence and Games (CIG '07)|
|Period||1/04/07 → 5/04/07|
- artificial intelligence
- video games
- games technology
Thompson, T., Levine, J., & Hayes, G. (2007). EvoTanks: co-evolutionary development of game-playing agents. Paper presented at IEEE 2007 Symposium on Computational Intelligence and Games (CIG '07), Hawaii, USA, .