Open AccessZnaczenie doboru metryk w badaniu separacji między klastrami
Author(s) -
Łukasz Paśko,
Galina Setlak
Publication year - 2016
Publication title -
studia informatica
Language(s) - English
Resource type - Journals
ISSN - 1731-2264
DOI - 10.21936/si2016_v37.n1.753
Subject(s) - separation (statistics) , selection (genetic algorithm) , metric (unit) , cluster (spacecraft) , space (punctuation) , feature selection , computer science , data mining , mathematics , artificial intelligence , machine learning , engineering , programming language , operating system , operations management
The aim of this paper is to examine the importance of selection of metric during the analysis of separation between clusters of objects in the feature space. Fourteen metrics known from the literature were selected for the calculations. Seven datasets that differ in the number of objects, attributes, and clusters were examined. For each of them, the four cluster separation measures were calculated. The article contains selected results with particular emphasis on the differences arising from the use of various metrics.
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom