Combination of Multivariate Curve Resolution with Factorial Discriminant Analysis for detection of grapevine diseases using Hyperspectral imaging. Case study: Flavescence Dorée

Hyperspectral imaging is an emergent technique in viticulture that can potentially detect viral diseases in a non-destructive manner. However, the main problem is to handle the huge amount of information...

Multi-Partitions Subspace Clustering

In model based clustering, it is often supposed that only one clustering latent variable explains the heterogeneity of the whole dataset. However, in many cases several latent variables could explain the heterogeneity of the data at hand. Finding such class variables could result in a richer interpretation of the data. In the continuous data setting, a multi-partition model based clustering is proposed. It assumes the existence of several latent clustering variables, each one explaining the heterogeneity of the data with respect to some clustering subspace. It allows to simultaneously find the multi-partitions and the related subspaces. Parameters of the model are estimated through an EM algorithm relying on a probabilistic reinterpretation of the factorial discriminant analysis. A model choice strategy relying on the BIC criterion is proposed to select to number of subspaces and the number of clusters by subspace. The obtained results are thus several projections of the data, each one conveying its own clustering of the data. Model’s behavior is illustrated on simulated and real data.

Classification of sugarcane varieties using visible/near infrared spectral reflectance of stalks and multivariate methods

The use of fast and non-destructive techniques for identifying sugarcane varieties enables the development of automatic sorting systems, contributing towards improving pre-processing steps in the alcohol and sugar industries. In this context, principal component analysis (PCA), factorial discriminant analysis (FDA), stepwise forward discriminant analysis (SFDA) and partial least-squares discriminant analysis (PLS-DA) were used to classify four Brazilian sugarcane varieties based on visible/near infrared (Vis/NIR) spectral reflectance measurements (450–1000 nm range) of stalks. All wavelengths contributed towards discriminating the sugarcane varieties, but the 600–750 nm range was most relevant. When evaluating PCA results considering the four sugarcane varieties, two of them overlapped and it was only possible to use classifiers of three varieties. Factorial discriminant analysis, PLS-DA and SFDA reached correct classifications of 0.81, 0.82 and 0.74, respectively, when considering the external validation data and the four sugarcane varieties evaluated. Results showed that Vis/NIR spectroscopy combined with discriminating methods is a promising tool for non-destructive and fast sugarcane variety classification, which can be used in the agro-food industry or directly in the field.

Saccharomyces cerevisiaeMixed Culture of Blackberry (Rubus ulmifoliusL.) Juice: Synergism in the Aroma Compounds Production

Blackberry (Rubussp.) juice was fermented using four different strains ofSaccharomyces cerevisiae(Vitilevure-CM4457, Enoferm-T306, ICV-K1, and Greroche Rhona-L3574) recognized because of their use in the wine industry. A medium alcoholic graduation spirit (<6GL°) with potential to be produced at an industrial scale was obtained. Alcoholic fermentations were performed at28C°, 200 rpm, and noncontrolled pH. The synergistic effect on the aromatic compounds production during fermentation in mixed culture was compared with those obtained by monoculture and physic mixture of spirits produced in monoculture. The aromatic composition was determined by HS-SPME-GC. The differences in aromatic profile principally rely on the proportions in aromatic compounds and not on the number of those compounds. The multivariance analysis, principal component analysis (PCA), and factorial discriminant analysis (DFA) permit to demonstrate the synergism between the strains.

SYMBOLIC FACTORIAL DISCRIMINANT ANALYSIS FOR ILLUMINATION INVARIANT FACE RECOGNITION

In this paper, a new appearance-based technique called symbolic factorial discriminant analysis (symbolic FDA) is explored for face representation and recognition under varying illumination conditions. In the past few years, many appearance-based methods have been proposed to model image variations of human faces under different lighting conditions using single valued variables to represent the facial features. In the proposed symbolic factorial discriminant analysis method, we extract interval type discriminating features, which are robust to illumination changes. The minimum distance classifier with symbolic dissimilarity measure is used for classification. The proposed method has been successfully tested for face recognition using three databases, namely, Yale Face database B, CMU PIE database and Harvard database. The experimental results have demonstrated the effective performance of this method.