Prototype based models offer an intuitive interface to given data sets
by means of an inspection of the model prototypes. Many of these
methods are however limited to the analysis of Euclidean vectors
and can not be used if data are characterized by general (dis-)
similarities such as problems specific measures.
In this talk we provide an overview on relational extensions
of modern prototype methods. We review applications for clustering,
classification and visualization employing similarity data.
back to the list of talks