TRATAMENTO GRÁFICO E MATEMÁTICO-ESTATÍSTICO DE DADOS MULTIVARIADOS: ANÁLISES DE AGRUPAMENTOS A PARTIR DAS TÉCNICAS MATRIZ ORDENÁVEL, COMPONENTES PRINCIPAIS, WARD E K-MÉDIAS

Auteurs

  • Márcia M. Duarte dos Santos Depto. de Geografia, IGC, UFMG
  • Miguel C. Sanchez Depto. de Estatística, ICEX, UFMG
  • Sueli Ap. Mingoti Depto. de Planejamento, IGCE, UNESP

DOI :

https://doi.org/10.18285/geonomos.v6i2.173

Résumé

In this paper some methods commonly used in multivariate cluster analysis are discussed and compared byusing a specific data set. The main objective is to show that the graphical method is efficient and similar tothe Ward and K-Means methods which are based upon mathematical and statistical theories. Thesimilarityfound between the graphical and the statistical methods suggests that although it is more subjective thegraphical method is a valid technique that could be used to determine the clusters of a sample or a population.,

Téléchargements

Les données relatives au téléchargement ne sont pas encore disponibles.

Téléchargements

Publiée

1998-12-01

Numéro

Rubrique

Artigos