Swarm art with KANTS: Using an ant clustering algorithm for generating abstract paintings

摘要

This paper describes an ant-based clustering algorithms and suggests its application as a swarm art conceptual tool. The algorithm, called KANTS, consists on a set of equations that model the local behavior of simple units (ants) that represent data samples. When moving on a heterogeneous 2-dimensional lattice of vectors, global behavior emerges from the local interaction between the ants, which self-organize into clusters. In this paper, KANTS is used outside a purely scientific framework. The algorithm is applied to two different types of data (sleep Electroencephalogram signals and colored photographs) for generating 2-dimensional abstract representations of those sets of data. We call the images pherogenic drawings, since they are a visual representation of the algorithm's pheromone maps. The method is contextualized within the swarm art field as well as within the authors' previous works on swarm art.