Principal Components Analysis with Optimal Scaling: A way to integrate sensory analysis, physico-chemical measurements and experimental factors.
2024
Vigneau, Evelyne | Bougeard, Stéphanie | École nationale vétérinaire, agroalimentaire et de l'alimentation Nantes-Atlantique (ONIRIS) | Statistique, Sensométrie et Chimiométrie (StatSC) ; École nationale vétérinaire, agroalimentaire et de l'alimentation Nantes-Atlantique (ONIRIS)-Institut National de Recherche pour l’Agriculture, l’Alimentation et l’Environnement (INRAE) | Laboratoire de Ploufragan-Plouzané-Niort [ANSES] ; Agence nationale de sécurité sanitaire de l'alimentation, de l'environnement et du travail (ANSES) | Société Française de Statistique (SFdS) & Polytechnic Institute of Bragança (IPB)
International audience
Afficher plus [+] Moins [-]anglais. Most multivariate statistical methods have been developed for quantitative/numerical data, while it is often the case that qualitative variables are collected at the same time. These qualitative/categorical variables can be either nominal, with a finite number of modalities identified by labels, or ordinal, where the modalities are ordered. To deal with mixed types of variables, an attractive strategy is to assign numerical values to the modalities. This operation of quantification or scaling can be carried out to optimize a criterion related to the objective of the data analysis, i.e. supervised or unsupervised. For an overview of the various approaches to optimal scaling that have been proposed in the past, see Abdi et al. (2024). In this presentation, we consider an exploratory analysis of the relationships between variables of mixed types, more specifically PCA-OS (Principal Components Analysis with Optimal Scaling), also known as the PRINQUAL procedure in SAS software.
Afficher plus [+] Moins [-]Mots clés AGROVOC
Informations bibliographiques
Cette notice bibliographique a été fournie par Institut national de la recherche agronomique
Découvrez la collection de ce fournisseur de données dans AGRIS