Exploring Interaction, Diversity and Efficiency of Biologically Inspired Evolutionary Multiagent Systems
2: University of Liverpool
This short paper presents work exploring properties of an evolutionary multiagent system assigned to solve sequential task achievement problems in dynamic, real-time, asynchronous environments. Several evolutionary models have been implemented and experiments conducted in a high-performance computing environment in which different interaction mechanisms and population diversity modes are evaluated according to multiple performance metrics. Results are presented that illustrate differences in performance efficiency when different interaction mechanisms, population diversity and evolutionary models are employed.
- Published: 2nd Feb 2015
- Publisher: ACM