Comparison Analysis of Clustering Quality Criteria Using Inductive Methods of Objective Clustering

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F44555601%3A13440%2F20%3A43895694" target="_blank" >RIV/44555601:13440/20:43895694 - isvavai.cz</a>

Výsledek na webu

<a href="https://link.springer.com/book/10.1007/978-3-030-61656-4" target="_blank" >https://link.springer.com/book/10.1007/978-3-030-61656-4</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Comparison Analysis of Clustering Quality Criteria Using Inductive Methods of Objective Clustering

Popis výsledku v původním jazyce

In this paper, we present the results of the research concerning comparison analysis of both the internal and external clustering quality criteria for clustering various types of datasets using density-based DBSCAN clustering algorithm implemented based on Inductive Methods of Objective Clustering (IMOC). Implementation of the IMOC technique assumes division of the initial dataset into two similar subsets contained the same number of pairwise similar objects at the first step of this procedure implementation. Then, we have executed the data clustering on the obtained subsets concurrently within the range of the appropriate algorithm parameters variation with estimation of various types of clustering quality criteria (internal (IQC) and external (EQC)) at each step of this procedure implementation. The final solution concerning algorithm optimal parameters determination was made based on the maximum values of the complex balance criterion (CBC) which contains both the ICQ and ECQ as the components. The analysis of the simulation results has allowed us to evaluate the effectiveness of both the internal and external clustering quality criteria to determine the optimal parameters of clustering algorithm using various type of data. To our mind, the obtained results can allow us to increase the clustering procedure exactness and to decrease the reproducibility error.

Název v anglickém jazyce

Comparison Analysis of Clustering Quality Criteria Using Inductive Methods of Objective Clustering

Popis výsledku anglicky

In this paper, we present the results of the research concerning comparison analysis of both the internal and external clustering quality criteria for clustering various types of datasets using density-based DBSCAN clustering algorithm implemented based on Inductive Methods of Objective Clustering (IMOC). Implementation of the IMOC technique assumes division of the initial dataset into two similar subsets contained the same number of pairwise similar objects at the first step of this procedure implementation. Then, we have executed the data clustering on the obtained subsets concurrently within the range of the appropriate algorithm parameters variation with estimation of various types of clustering quality criteria (internal (IQC) and external (EQC)) at each step of this procedure implementation. The final solution concerning algorithm optimal parameters determination was made based on the maximum values of the complex balance criterion (CBC) which contains both the ICQ and ECQ as the components. The analysis of the simulation results has allowed us to evaluate the effectiveness of both the internal and external clustering quality criteria to determine the optimal parameters of clustering algorithm using various type of data. To our mind, the obtained results can allow us to increase the clustering procedure exactness and to decrease the reproducibility error.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2020

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Communications in Computer and Information Science

ISBN

978-3-030-61655-7

ISSN

1865-0929

e-ISSN

1865-0937

Počet stran výsledku

17

Strana od-do

150-166

Název nakladatele

Springer Nature Switzerland AG

Místo vydání

Switzerland

Místo konání akce

Lviv, Ukraine

Datum konání akce

21. 8. 2020

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—