Clustering of Variables Around Latent Components

Analiza Zasobów Kompetencji Z Wykorzystaniem Grupowania Zmiennych Niemetrycznych

“…Jak wskazano w opracowaniu [Chavent i in. 2013] do najważniejszych metod dla zmiennych metrycznych należy zaliczyć: procedurę VARCLUS dostępną w oprogramowaniu SAS, metodę grupowania wokół zmiennych ukrytych [Vigneau, Qannari 2003, Vigneau i in. 2015], diametrical clustering [Dhillon i in.…”

Missing Data and Imputation Methods in Partition of Variables

“…We optimize S rather than a criterion based on squared correlation because, for us, in many situations, the sign of the correlation coefficient makes sense, see Vigneau and Qannari(2003): " p consumers are asked to rate their acceptability of n product. A negative covariance between the scores of two consumers emphasizes their different views of the products.…”

“…Our method 5 is derived from Vigneau and Qannari(2003) and closely related to principal component analysis. We may also say that this approach consists in clustering variables around latent components.…”

“…This hierarchical cluster method is based on the same criterion S described for the partition see Vigneau and Qannari(2003).…”

ClustOfVar: AnRPackage for the Clustering of Variables

“…The most famous one is the VARCLUS procedure of the SAS software (SAS Institute Inc. 2011). Recently specific methods based on PCA were proposed by Vigneau and Qannari (2003) with the name clustering around latent variables (CLV) and by Dhillon, Marcotte, and Roshan (2003) with the name Diametrical Clustering. But all these specific approaches work only with quantitative data and as far as we know, they are not implemented in R.…”