The aim of the present paper is to reconsider and re-interpret the iterative factor clustering for binary data in a two-fold way: a) define the dimension reduction step in the procedure as a statistical model based on a probability distribution; b) relax the hard classification (nonoverlapping) produced by the K-means, and allow for partial classification of respondents, that can be assigned to multiple clusters, with a different degree of membership.
Iterative factor clustering of categorical data reconsidered
IODICE D'ENZA, Alfonso;
2015-01-01
Abstract
The aim of the present paper is to reconsider and re-interpret the iterative factor clustering for binary data in a two-fold way: a) define the dimension reduction step in the procedure as a statistical model based on a probability distribution; b) relax the hard classification (nonoverlapping) produced by the K-means, and allow for partial classification of respondents, that can be assigned to multiple clusters, with a different degree of membership.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
