In this paper, two evolutionary algorithms for clustering in the domain of directed weighted graphs are proposed. Several genetic operators are analyzed with respect to maintaining the balance between exploration and exploitation properties. The approach is extensively tested on medium-sized random graphs.
展开▼