K.A. Hawick, H.A. James, and C.J. Scogings (New Zealand)
classification; clustering; chaotic and complex systems; multi-agent systems.
We discuss algorithms and methods for classifying the clusters of model animals that emerge from simula tions of collective behaviour in artificial life models. We show how important statistical properties for understand ing scaling and universal growth can be measured from complex and chaotic model systems. We describe animal clustering algorithms and the difficulties involved in auto matic tracking of herds that move and change shape, orien tation and size in time. We present some heuristic rules for semi-automated classification over time and some prelimi nary results from our study of a predator-prey multi-agent model.
Important Links:
Go Back
