Abstract:
In a game it is often the case that there are multiple roles or types of actors with different goals. One possible target for automatic content generation is to create mu...Show MoreMetadata
Abstract:
In a game it is often the case that there are multiple roles or types of actors with different goals. One possible target for automatic content generation is to create multiple different software agents for these distinct roles. This paper outlines a technique, based on the multiple worlds model, for creating such actors via evolution. The objective function is based on the performance of the actors within their role and retains the ability of evolution to operate on populations of agents, permitting the creation of many possible agents for each role. The fitness of distinct agent types is never compared by this algorithm, so the evolution of agents affects fitness only at the level of the user specified simulation used to evaluate performance. The technique is demonstrated by simultaneously evolving four populations of prisoner's dilemma agents with very different roles.
Date of Conference: 22-25 August 2017
Date Added to IEEE Xplore: 26 October 2017
ISBN Information:
Electronic ISSN: 2325-4289