Modeling hierarchical and modular complex networks based on FCM

Abstract

In this paper, we investigate the construction of complex networks based on clustering idea. Firstly, the resulting networks are woven by the clustering paths which follow their cluster's "centroids". Secondly, when the number of the data is huge, the data will be divided into subsets at different levels according to their similarity. The presented algorithm will be carried out in, between, and among these subsets at different levels. The resulting networks display small world feature and community structure, characterized by the hierarchical clustering function of a vertex with degree k, c(k) like some real-world networks. We also study the evolution behaviors and formation mechanism of the resulting networks. Index Terms—complex networks, scale-free, small world, fuzzy c-means .