Using Singularity Exponent in Distance based Classifier

Autoři: Marcel Jiřina
Rok vydání: 2010
Typ publikace: Konferenční příspěvek (mezinárodní konf.)
Periodikum: Proceedings of the 2010 10th International Conference on Intelligent Systems Design and Applications
Projekty: Centrum aplikované kybernetiky
Bibliografická citace: JIŘINA, Marcel, JIŘINA JR., M. Using Singularity Exponent in Distance based Classifier. In: HASSANIEN, A.E., ABRAHAM, A., MARCELLONI, F., HAGRAS, H., ANTONELLI, M., HONG, T.P., eds. Proceedings of the 2010 10th International Conference on Intelligent Systems Design and Applications. Los Alamitos: IEEE, 2010, S. 220-224. ISBN 978-1-4244-8135-4
Dostupné z: doi:10.1109/ISDA.2010.5687263
DOI: 10.1109/ISDA.2010.5687263
ID: 351724

Abstrakt

The paper deals with using so called singularity exponent in a classifier that is based on ordered distances of patterns to a given (classified) pattern. The approximation of probability distribution mapping function of the distribution of points from the viewpoint of distances from a given point in a form of a suitable power (exponent) of a distance is presented together with a way how to state it. A classifier utilizing knowledge about explored data distribution in a space and a suggested expression of the exponent is presented. Experimental results on both synthetic and real-life data show interesting behavior (classification accuracy) of the classifier in comparison with other well-known classifiers.