IC: Intelligent Clustering, a new time efficient data partitioning methodology

Stojanov, Done (2012) IC: Intelligent Clustering, a new time efficient data partitioning methodology. International Journal of Computer Science and Information Technologies, 3 (5). pp. 5065-5067. ISSN 0975-9646

[img]
Preview
Text
__ugd.edu.mk_private_UserFiles_done.stojanov_Desktop_ijcsit2012030525.pdf

Download (619Kb) | Preview
Official URL: http://www.ijcsit.com/docs/Volume%203/vol3issue5/i...

Abstract

A new, time efficient data partitioning methodology is presented. First, the input data set is mapped into a set of radius vectors, which is further sorted, then partitioned into initial clusters. However, that is not a guarantee that all objects have been partitioned in the appropriate clusters. Therefore, objects being inappropriately partitioned are moved from one cluster into another neighbouring cluster. Finally, clusters of radius vectors are formed. Applying reverse mapping, clusters of radius vectors are back-mapped into clusters of original objects, being appropriately partitioned.

Item Type: Article
Subjects: Natural sciences > Computer and information sciences
Divisions: Faculty of Computer Science
Depositing User: Done Stojanov
Date Deposited: 31 Oct 2012 18:52
Last Modified: 31 Oct 2012 18:52
URI: http://eprints.ugd.edu.mk/id/eprint/189

Actions (login required)

View Item View Item