Community Detection in Complex Networks by Dynamical Simplex Evolution

TL;DR

This paper evaluates the dynamical simplex evolution (DSE) method for community detection in complex networks, comparing its accuracy with other algorithms and exploring its ability to identify hierarchical structures.

Contribution

It introduces and benchmarks the DSE method against existing algorithms for community detection, highlighting its effectiveness and potential for hierarchical analysis.

Findings

DSE performs well in correctly identifying nodes across various network fuzziness levels.

DSE shows promise in detecting hierarchical substructures within complex networks.

Benchmark results favor DSE in certain network configurations.

Abstract

We benchmark the dynamical simplex evolution (DSE) method with several of the currently available algorithms to detect communities in complex networks by comparing the fraction of correctly identified nodes for different levels of ``fuzziness'' of random networks composed of well defined communities. The potential benefits of the DSE method to detect hierarchical sub structures in complex networks are discussed.