Evolving agent swarms for clustering and sorting

Social insect societies consist of individuals with simple behavior, yet at the collective level the societies are capable of solving complex tasks. These tasks span a wide range, including clustering, patch sorting, and annular sorting. It has proved particularly difficult to recreate the social insects ability to perform annular sorting, even for engineered solutions. In this paper we present the results we have obtained from evolving swarms of agents that are capable of performing clustering, patch sorting, and annular sorting. Most noteworthy is our solution to the 3-type annular sorting problem which has previously not been solved successfully. Categories and Subject Descriptors: I.2.m [Miscellaneous] General Terms: Experimentation

• Type: Conference
• Year: 2005
• Where: GECCO
• Authors: Vegard Hartmann
• Keywords: Annular Sorting Problem, GECCO 2005, Social Insect Societies, Social Insects