In the Food research and production field, system complexity is increasing and several new challenges are emerging every day. This implies a urgent necessity to extract information and obtain models capable of inferring the underlying relationships that link all the variability sources which characterize food or its production process (e.g. compositional profile, processing conditions) to very general end-properties of foodstuff, such as the healthiness, the consumer perception, the link to a territory and the effect of the production chain itself on food.\ud This makes a “deductive”, theory-driven research approach inefficient, since it is often difficult to formulate hypotheses. Explorative Multivariate Data Analysis methods, together with the most recent analytical instrumentation, offer the possibility to come back to an “inductive” data-driven attitude with a minimum of a priori hypotheses, instead helping in formulating new ones from the direct observation of data.\ud The aim of this Chapter is to offer the reader an overview of the most significant tools which can be used in a preliminary, exploratory phase, ranging from the most classical descriptive statistics methods, to Multivariate Analysis methods, with particular attention to Projection methods. For all techniques, examples are given so that the main advantage of this techniques, that is a direct, graphical representation of data and their characteristics, can be immediately experienced by the reader

Multilinear PLS (NPLS) and its discriminant version (NPLS-DA) are very diffuse tools to model multi-way data\ud arrays. Analysis of NPLS weights and NPLS regression coefficients allows data patterns, feature correlation\ud and covariance structure to be depicted. In this study we propose an extension of the Variable Importance\ud in Projection (VIP) parameter to multi-way arrays in order to highlight the most relevant features to predict\ud the studied dependent properties either for interpretative purposes or to operate feature selection. The VIPs\ud are implemented for each mode of the data array and in the case of multivariate dependent responses considering\ud both the cases of expressing VIP with respect to each single y-variable and of taking into account\ud all y-variables altogether.\ud Three different applications to real data are presented: i) NPLS has been used to model the properties of\ud bread loaves from near infrared spectra of dough, acquired at different leavening times, and corresponding\ud to different flour formulations. VIP values were used to assess the spectral regions mainly involved in determining\ud flour performance; ii) assessing the authenticity of extra virgin olive oils by NPLS-DA elaboration of\ud gas chromatography/mass spectrometry data (GC–MS). VIP values were used to assess both GC and MS discriminant\ud features; iii) NPLS analysis of a fMRI-BOLD experiment based on a pain paradigm of acute\ud prolonged pain in healthy volunteers, in order to reproduce efficiently the corresponding psychophysical\ud pain profiles. VIP values were used to identify the brain regions mainly involved in determining the pain intensity\ud profile

Nowadays the necessity to reveal the hidden information from complex data sets is increasing due to the development of high-throughput instrumentation. The possibility to jointly analyze data sets arising from different sources (e.g. different analytical determinations/platforms) allows capturing the latent information that would not be extracted by the individual analysis of each block of data. Several approaches are proposed in the literature and are generally referred to as data fusion approaches. In this work a mid level data fusion is proposed for the characterization of three varieties (Salamino di Santa Croce, Grasparossa di Castelvetro, Sorbara) of Lambrusco wine, a typical PDO wine of the district of Modena (Italy). Wine samples of the three different varieties were analyzed by means of 1H-NMR spectroscopy, Emission-Excitation Fluorescence Spectroscopy and HPLC-DAD of the phenolic compounds.\ud \ud Since the analytical outputs are characterized by different dimensionalities (matrix and tensor), several multivariate analyses were applied (PCA, PARAFAC, MCR-ALS) in order to extract and merge, in a hierarchical way, the information present in each data set.\ud \ud The results showed that this approach was able to well characterize Lambrusco samples giving also the possibility to understand the correlation between the sources of information arising from the three analytical techniques

In the last decades, mankind has become totally aware about the importance of food quality: nowadays authentication and traceability are words of general use. Food authentication verifies how much a food is in accordance with its label description and law and it could be considered a further guarantee for the quality and safety of a foodstuff. The traceability of food could be considered an essential element in ensuring safety and high quality of food. The synergistic use of instrumental analytical techniques and chemometrics represents a promising way to obtain trustworthy results in the development of authenticity and traceability models. This chapter deals with the potentialities of chemometrics tools in resolving some real issues related to food traceability and authenticity. Particular attention will be paid to the use of some exploratory, classification and discrimination techniques. In the first part of this chapter, a briefly description of European regulations (Authenticity and Traceability: the European Union point of view), and traceability and authenticity markers (Authenticity and Traceability: a scientific point of view) is reported. The second part is split into two sections: namely Food Authenticity and Food Traceability applications, where the main features and advantages of some chemometrics approaches are presented.

No abstract