Clustering datasets by means of CLUSTATIS with identification of atypical datasets. Application to sensometrics

Abstract A method of clustering a collection of datasets measured on the same individuals, called CLUSTATIS, was introduced and applied to the segmentation of the subjects participating in a projective mapping experiment or a free sorting task. A refinement of this method of clustering is proposed. It consists in segmenting the subjects while discarding those subjects who can be considered as atypical because they do not fit to the pattern of any cluster. This strategy of analysis requires the determination of a threshold parameter that delineates the boundary between the main clusters and the noise cluster that contains the atypical subjects. An appropriate selection of this parameter is proposed. The general strategy of analysis is illustrated on the basis of a simulation study and data from projective mapping and free sorting experiments.

[1]  Fabien Llobell,et al.  Analysis and clustering of multiblock datasets by means of the STATIS and CLUSTATIS methods. Application to sensometrics , 2020 .

[2]  Tormod Næs,et al.  Estimating and interpreting more than two consensus components in projective mapping: INDSCAL vs. multiple factor analysis (MFA) , 2017 .

[3]  François-Joseph Lapointe,et al.  Using the stability of objects to determine the number of clusters in datasets , 2017, Inf. Sci..

[4]  Gastón Ares,et al.  Influence of consumers' cognitive style on results from projective mapping. , 2017, Food research international.

[5]  Rajesh N. Davé,et al.  Characterization and detection of noise in clustering , 1991, Pattern Recognit. Lett..

[6]  Frank Westad,et al.  Strategies for consumer segmentation with applications on preference data , 2004 .

[7]  Tormod Næs,et al.  Segmentation in projective mapping , 2019, Food Quality and Preference.

[8]  Evelyne Vigneau,et al.  Segmentation of consumers in preference studies while setting aside atypical or irrelevant consumers , 2016 .

[9]  P. Robert,et al.  A Unifying Tool for Linear Multivariate Statistical Methods: The RV‐Coefficient , 1976 .

[10]  Sébastien Lê,et al.  SENSOMINER : A PACKAGE FOR SENSORY DATA ANALYSIS , 2008 .

[11]  Pascal Schlich,et al.  Defining and Validating Assessor Compromises About Product Distances and Attribute Correlations , 1996 .

[12]  Pascal Schlich,et al.  SORT-CC: A procedure for the statistical treatment of free sorting data , 2010 .

[13]  Robert Sabatier,et al.  The ACT (STATIS method) , 1994 .

[14]  El Mostafa Qannari,et al.  Measures of association between two datasets; Application to sensory data , 2015 .