Alessandra Biancolillo defended her PhD thesis at the University of Copenhagen on November 18th. Alessandra has been associated with Nofima’s strategic program “Multiblock Methods for prediction and interpretation”, developing methodology for analyzing and interpreting several blocks of data sequentially (so-called SO-PLS). The thesis is titled “Method development in the area of multi-block analysis focused on food analysis”. Supervisors have been Tormod Næs (Nofima), Ingrid Måge (Nofima) and Rasmus Bro (University of Copenhagen).
Alessandra did a very good job at the defense. Opponents Thomas Skov (University of Copenhagen), El Mostafa Qannari (Oniris, Nantes) and Lars Nørgaard (Roskilde University) were well prepared and contributed to an interesting and fruitful discussion.
The 14’th Scandinavian Symposium in Chemometrics (SSC14) were held in Chia in Sardinia June 14-17. Some may think this was outside Scandinavia, and although the island has a nice coast and a fantastic nature, they are right. The advantageous of having a conference in Sardinia is, however, nice temparatures, nice beaches, nice temperatures for swimming and last, but at not least nice food and drinks. There were approximately 140 participants.
Participants from Nofima and NMBU were
Kristian Hovde Liland: “Comparing different measurement technologies by considering low-dimensional projections of associated sample data matrices”
Ingunn Berget: «Critical comparison of distance based methods, PCA and multiblock analyses of Polarized Sensory Positioning data” (flash presentation)
Ulf Indahl (NMBU): “Explaining the secrets behind O-PLS and Target Projections to non-experts by elementary linear algebra.”
The paper entitled “Modeling target group heterogeneity in experimental consumer studies” coauthored by J. Jansen, E. Menichelli and T. Næs was recently published in Food Quality and Preference. The paper presents the method Indvidual Differences (InD) to deal with heterogeneity in consumer groups.
Jansen, J. J., Menichelli, E., & Næs, T. (2015). Modeling target group heterogeneity in experimental consumer studies. Food Quality and Preference, 45(0), 50-57.
Consumer group heterogeneity, Consumer segmentation ,Individual Differences (InD) method, ANOVA residuals
Acceptance of a product by a consumer may result from a convoluted interplay between product attributes and individual characteristics of that consumer. Different methods that systematically combine product properties with consumer groups segmented on such characteristics have provided unprecedented insight, but ignore heterogeneity in acceptance within each consumer group. Although such knowledge is invaluable for targeted marketing, dedicated methods for consumer group heterogeneity are lacking. The authors aim to fill this gap by the Individual Differences (InD) method, which models differences between consumers within the same target group. The method scores the ‘diffusion’ within each group, shows how much each consumer contributes to that, and relates this information to product properties. Thereby also novel groups may be discovered, with attributes not covered in the consumer segmentation. The illustrative consumer study on apple juice reveals how young women differ in their price-consciousness and their acceptance on specific preparation technologies more than older women. Although men exhibit heterogeneity on the same product attributes, their mutual variability is considerably lower and they thereby form more homogeneous target groups.
The paper “A comparison of generalised procrustes analysis and multiple factor analysis for projective mapping data” authored by O. Tomic, I. Berget and T. Næs was recently published in Food Quality and Preference.
Generalised procrustes analysis and multiple factor analysis are multivariate statistical methods that
belong to the family of multiblock methods. Both methods are often used for analysis of data from projective
mapping (a.k.a. Napping). In this study, generalised procrustes analysis and multiple factor analysis
are compared for a number of simulated and real data sets. The type of data used in this study
were (I) random data from Monte Carlo simulations; (II) constructed data that were manipulated according
to some specific criteria; (III) real data from nine Napping experiments. Focus will be on similarities of
the consensus solutions. In addition we considered interpretation of the RV coefficient and individual differences
Generalised procrustes analysis
Multiple factor analysis
Combining SO-PLS and linear discriminant analysis for multi-block classification
The “Combining SO-PLS and linear discriminant analysis for multi-block classification”, written by Alessandra Biancolillo, Ingrid Måge and Tormod Næs was recently published in Chemometrics and Intelligent Laboratory systems. This is the first paper by Ph.D student Alessandra. Congratulations!
The aim of the present work is to extend the Sequentially Orthogonalized-Partial Least Squares (SO-PLS) regression method, usually used for continuous output, to situations where classification is the main purpose. For this reason SO-PLS discriminant analysis will be compared with other commonly used techniques such as Partial Least Squares-Discriminant Analysis (PLS-DA) and Multiblock-Partial Least Squares Discriminant Analysis (MB-PLS-DA). In particular we will focus on how multiblock strategies can give better discrimination than by analyzing the individual blocks. We will also show that SO-PLS discriminant analysis yields some valuable interpretation tools that give additional insight into the data. We will introduce some new ways to represent the information, taking into account both interpretation and predictive aspects.
Biancolillo, A., Måge, I., & Næs, T. (2015). Combining SO-PLS and linear discriminant analysis for multi-block classification. Chemometrics and Intelligent Laboratory Systems, 141(0), 58-67.
The software used in this paper, can be downloaded from software&downloads
Paula Varela has been coauthoring the paper “Understanding consumers’ perception of the concept and sensory experience of a functional food” to be published in LWT-Food Science and Technology. In this paper the authors used Flash Profile to generate attributes to further use in a CATA study with consumers. Sensory and non-sensory parameters related to food choice were explored. The full reference of the paper with link to ScienceDirect is
Hernández-Carrión, M., Varela, P., Hernando, I., Fiszman, S. M., & Quiles, A. Persimmon milkshakes with enhanced functionality: Understanding consumers’ perception of the concept and sensory experience of a functional food. LWT – Food Science and Technology.
Highlights from paper
Consumers perceived persimmon milkshakes as a functional beverage.
Milkshakes with pressurised persimmon were highly accepted by consumers.
Milkshakes with pasteurised persimmon were scored with the lowest overall liking.
Pressurisation allowed formulating persimmon milkshakes despite their seasonality.