Classification using fuzzy geometric features

The work describes a new approach to multidimensional data classification or clustering. "Radar" diagrams are the well-known and most widely used data visualization technique with an ability to be used in expert and man-machine systems. This paper extends our earlier work (Karkishchenko et al., 2000) in image analysis and classification by means of fuzzy geometric features for the fuzzy classification of arbitrary data, presented as a radar diagram. We develop a diagram shape matching methodology to accomplish the fuzzy geometric features technique for the man-machine expert systems.