An efficient fractals-based algorithm for clustering

Fractals, sets that exhibit certain self-similarity, have attracted considerable attention in the past decade. They have been used to model a wide variety of natural phenomena. This paper presents a new algorithm based on fractals for clustering. In particular, an algorithm to cluster 2D data based on the correlation dimension is presented. The algorithm is simple to implement and has low computational complexity. Experiments applying the algorithm to different datasets are presented and confirm the suitability of the approach for this application.