Self-organizing map of complex networks for community detection
- 279 Downloads
Detecting communities from complex networks is an important issue and has attracted attention of researchers in many fields. It is relevant to social tasks, biological inquiries, and technological problems since various networks exist in these systems. This paper proposes a new self-organizing map (SOM) based approach to community detection. By adopting a new operation and a new weight-updating scheme, a complex network can be organized into dense subgraphs according to the topological connection of each node by the SOM algorithm. Extensive numerical experiments show that the performance of the SOM algorithm is good. It can identify communities more accurately than existing methods. This method can be used to detect communities not only in undirected networks, but also in directed networks and bipartite networks.
Key wordsCommunity detection complex network neural networks self-organizing map
Unable to display preview. Download preview PDF.
- L. Chen, R. S. Wang, and X. S. Zhang, Biomolecular Networks: Methods and Applications in Systems Biology, John Wiley and Sons, Hoboken, NJ, 2009.Google Scholar
- X. S. Zhang and R. S. Wang, Optimization analysis of modularity measures for network community detection, Lect. Notes Oper. Res., 2008, 9: 13–20.Google Scholar
- X. S. Zhang, Neural Networks in Optimization, Kluwer Academic Publishers, 2000.Google Scholar
- A. Davis, B. B. Gardner, and M. R. Gardner, Deep South, University of Chicago Press, Chicago, 1941.Google Scholar