SUWAN: A supervised clustering algorithm with attributed networks

An increasing area of study for economists and sociologists is the varying organizational structures between business networks. The use network science makes it possible to identify determinants performance these In this work we look inter-firm performance. On one hand, a new method supervised clustering with attributed networks proposed, SUWAN, aim at obtaining class-uniform clusters turnover, while minimizing number clusters. This deals representative-based clustering, where set initial representatives randomly chosen. One innovative aspects SUWAN that algorithm can be accomplished through combination weights matrix distances nodes their attributes when defining As benchmark, Subgroup Discovery on data. focuses detecting subgroups described by specific patterns are interesting respect some target concept explaining features. other in order analyze impact network’s topology group’s performance, measures, group total turnover were exploited. proposed methodologies applied an inter-organizational network, EuroGroups Register, central register contains statistical information from European countries.