Supervised Pattern Recognition Procedures for Discrimination of Whiskeys from Gas Chromatography/Mass Spectrometry Congener Analysis

Abstract

The volatile congener analysis of 52 commercialized whiskeys (24 samples of single malt Scotch whiskey, 18 samples of bourbon whiskey, and 10 samples of Irish whiskey) was carried out by gas chromatography/mass spectrometry after liquid-liquid extraction with dichloromethane. Pattern recognition procedures were applied for discrimination of different whiskey categories. Multivariate data analysis includes linear discriminant analysis (LDA), k nearest neighbors (KNN), soft independent modeling of class analogy (SIMCA), procrustes discriminant analysis (PDA), and artificial neural networks techniques involving multilayer perceptrons (MLP) and probabilistic neural networks (PNN). Classification rules were validated by considering the number of false positives (FPs) and false negatives (FNs) of each class associated to the prediction set. Artificial neural networks led to the best results because of their intrinsic nonlinear features. Both techniques, MLP and PNN, gave zero FPs and zero FNs for all of the categories. KNN is a nonparametric method that also provides zero FPs and FNs for every class but only when selecting K = 3 neighbors. PDA produced good results also (zero FPs and FNs always) but only by selecting nine principal components for class modeling. LDA shows a lesser classification performance, because of the building of linear frontiers between classes that does not apply in many real situations. LDA led to one FP for bourbons and one FN for scotches. The worse results were obtained with SIMCA, which gave a higher number of FPs (five for both scotches and bourbons) and FNs (six for scotchs and two for bourbons). The possible cause of these findings is the strong influence of class inhomogeneities on the SIMCA performance. It is remarkable that in any case, all of the methodologies lead to zero FPs and FNs for the Irish whiskeys.

Quantitative chemical profile and multivariate statistical analysis of alembic distilled sugarcane spirit fractions

“…A data cross-validation was used to avoid the generation of an optimistic model and thus increase its prediction ability. 21,29 The calibration model to identify head, heart and tail fractions was built up with 21 samples being seven samples for each fraction. According to Table 1, this model presented efficient prediction ability for 90.5% of the samples and was validated with other nine different fractions whose origins were 100% correctly fitted.…”

Critical Aspects of Supervised Pattern Recognition Methods for Interpreting Compositional Data

“…Loadings of original features in the canonical variates and cumulative proportions of eigenvalues are interpreted, partly by analogy with PCA. Note that if one set of features are dummy variables giving group indicators, and then CVA is mathematically identical to LDA (González-Arjona et al, 2006). PLS-DA finds latent variables in the feature space which have a maximum covariance with the y variable.…”

“…These two concepts provide a deep knowledge of the classes' space. Accordingly, it seems to be more advisable the use of terms sensitivity (SENS) and specificity (SPEC) (González-Arjona et al, 2006) for validating the decision rule. The SENS of a class corresponds to the rate of evaluation objects belonging to the class that are correctly classified, and the SPEC of a class corresponds to the rate of evaluation objects not belonging to the class that are correctly considered as belonging to the other classes.…”

Optical spectroscopy and pattern recogition techniques for discriminating and classifying Scotch whiskies

“…can discriminate Scotch whiskies of different product categories [3][4]. For an instrumental assessment, authenticity analyses are carried out in specialized laboratories making use of gas chromatography, mass spectrometry, headspace solid-phase microextraction, and capillary electrophoresis [5][6][7]. Other olfactory/tasting classifications can be achieved by means of innovative e-noses and e-tongues [8][9].…”