Publica
Hier finden Sie wissenschaftliche Publikationen aus den FraunhoferInstituten. Exploiting the structures of the Umatrix
 Villmann, T.: Advances in SelfOrganizing Maps and Learning Vector Quantization : Proceedings of the 10th International Workshop, WSOM 2014, Mittweida, Germany, July, 24, 2014 Cham: Springer International Publishing, 2014 (Advances in Intelligent Systems and Computing 295) ISBN: 9783319076942 (Print) ISBN: 9783319076959 (Online) pp.249257 
 Workshop on Self Organizing Maps (WSOM) <10, 2014, Mittweida> 

 English 
 Conference Paper 
 Fraunhofer IME () 
Abstract
The Umatrix has become a standard visualization of selforganizing feature maps (SOM). Here we present the abstract Umatrix, which formalizes the structures on a Umatrix such that distance calculations between bestmatching units w.r.t. the height structures of a Umatrix are precisely defined (Ucell distance). This enables the assessment of the topological correctness of the SOM and the implementation of clustering algorithms that take the structures seen on the Umatrix into account. A weighted Delaunay graph of the Ucell distances allows the calculation of a dendrogram corresponding to the structures of the Umatrix. The method is shown to detect and visualize meaningful cluster structures on difficult artificial and reallife data.