by C. M. Fernandes, A. Mora, J.J. Merelo, V. Ramos, J.LJ. Laredo and A.C. Rosa
published in Proceedings of the 6th International Conference on Ant Colony Optimization and Swarm Intelligence (ANTS 2008), Lecture Notes in Computer Science LNSC 5217, Springer-Verlag, ISBN 978-3-540-87526-1, pp. 339-346, 2008.
Abstract. This paper investigates a new model that takes advantage of the cooperative self-organization of Ant Algorithms to evolve a naturally inspired pattern recognition (and also clustering) method. The approach considers each data item as an ant that changes the environment as it moves through it. The algorithm is successfully applied to well-known classification problems and yields better results than some other classification approaches, like K-Nearest Neighbours and Neural Networks.
Via communication through the environment (stigmergy), the ants/samples tend to form clusters, according to their class. Here, an example with the Iris data.
Carlos Miguel Fernandes
