Cluster Analysis is a well established methodology in Marketing research to perform market segmentation. Data from questionnaire surveys are often measured on ordinal scales, which are usually analyzed as they were of the interval type or by having recourse to non parametric distances like the \gamma index or distances based on the Spearman rank correlation. New dissimilarity distances dealing with variables of the ordinal type are presented.
Boari, G., Cantaluppi, G., Zanella, A., Some Distance Proposals for Cluster Analysis in Presence of Ordinal Variables, in Minerva, T., Morlini, I., Palumbo, F. (ed.), Cladag 2013. 9th Scientific Meeting of the Classification and Data Analysis Group of the Italian Statistical Society. Book of Abstracts, Cleup, Padova 2013: 49- 52 [http://hdl.handle.net/10807/50752]
Some Distance Proposals for Cluster Analysis in Presence of Ordinal Variables
Boari, Giuseppe;Cantaluppi, Gabriele;Zanella, Angelo
2013
Abstract
Cluster Analysis is a well established methodology in Marketing research to perform market segmentation. Data from questionnaire surveys are often measured on ordinal scales, which are usually analyzed as they were of the interval type or by having recourse to non parametric distances like the \gamma index or distances based on the Spearman rank correlation. New dissimilarity distances dealing with variables of the ordinal type are presented.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
