The spatial-temporary approach in problems of clusterization. Lyudmyla Sarycheva. IWIM, Prague, 2007.

 Article (in pdf)

Abstract. The problem of cluster analysis of the spatial-temporary data is considered. The new algorithm of GeoTime?-clusterization takes into account a temporary neighbourhood of objects features (on the basis of the inductive approach) and absolute concepts of objects location (on the basis of the geoinformation approach). Using the actual data of ecological and socio-economic monitoring of the Europe states the experimental comparison of GeoTime?-clusterization with known algorithms is carried out by three clusterization quality criteria and gave the good results.

Keywords. GeoTime?-clusterization, clusterization quality criteria, spatial-temporary data, socio-economic monitoring, proximity measures

References.

1. Classification and reduction of dimensionality. Edited by prof. S.A.Aivazyan, 1989. –608 p.
2. Tou J.T., Gonzalez R.C. Pattern recognition principles.– Addison-Wesley Publishing Company, 1974.– 411 p.
3. Duran B., Odell P. Cluster Analysis. – М.: Statistica, 1977. – 128 p.
4. Sarycheva L., Boyko A. Geomatic clusterization based on ecological-social-economical monitoring rates //Scientific Bulletin of NMU, 2006, №5.– P.76-81.
5. Eurostat // Эл. ресурс. URL:  http://www.eurostat.com/
6. Sarycheva L., Zhuykov A. Cluster Analysis of Territories by the Totality of Ecological and Socio-Economic Indices // IEEE 2001 International Geoscience and Remote Sensing Symposium, UNSW. – P.663-664.