B1382
Title: Clustering of star-shaped sets with a fuzzy approach
Authors: Maria Brigida Ferraro - Sapienza University of Rome (Italy) [presenting]
Elena Fernandez Iglesias - University of Oviedo, INDUROT (Spain)
Ana Belen Ramos-Guajardo - Fundacion Universidad de Oviedo (Spain)
Gil Gonzalez-Rodriguez - University of Oviedo (Spain)
Abstract: Star-shaped sets represent a large class of compact and convex sets as a particular case. The original shape of these sets is identified by means of the center-radial characterization. These data can be defined as complex because multiple aspects of them are enclosed in a single representation. Since, in some situations, it may be very useful to partition these sets into groups, a fuzzy clustering method is addressed. In detail, the proposal is a generalization of the well-known fuzzy $k$-means that takes into account the nature of the data through an appropriate distance measure. A real case study is reported to check the proposal's adequacy.
