Table 2.
Compilation of distinct analytical approaches combined with chemometrics for beer classification.
| Instrument | Technique | Goal | Sample | Accuracy* (%) | Ref |
|---|---|---|---|---|---|
| Gas chromatography-mass spectrometry | HS-SPME-GC-TOFMS and ANN-MLP | Discriminate trappist class and specific brands from non-trappist | 265 specialty beer samples | 93.9–97 | 37 |
| ISEs | Potentiometry and LDA | Discrimination of different commercial beer types | 51 different brands and varieties of beer | 81.9 | 26 |
| Fluorescence and UV–Vis spectrophotometer | Spectroscopy and PCA-LDA data fusion | classification of canned samples of Chinese lager beers by manufacturer | 135 canned beer samples from eleven Chinese manufacturers | 78.5–86.7 | 38 |
| Paper spray mass spectrometry | Paper spray mass spectrometry and OPS-PLS-DA | Differentiation of Brazilian American lager beers according to their brands | 141 samples from four breweries | 100 | 9 |
| Spectrometer | 1H NMR spectroscopy and PLSDA/SIMCA | Discriminate Standard and Premium Brazilian American lager beers | 20 Premium American Lager and 20 Standard American Lager | 91.6–100 | 23 |
| Fluorescence spectrophotometer | EEM fluorescence and PARAFAC-kNN | Characterization and classification of Chinese beers from different manufacturers | 108 canned beer samples from four major Chinese manufacturers | 91.7 | 39 |
| SPCE | Voltammetric and PLS-DA | Differentiation of Brazilian Premium american lager and Standard american lager | 59 Premium american lagers and 54 Standard american lagers | 94 | 14 |
| SPCE | Voltammetric and SVM-DA | Differentiation of Brazillian Beer at manufacturer and brand level | 253 beers from four major Brazilliam manufacturers | 96–98 | Present study |
*Accuracy: Rate of correct classification in relation to an external test set; SPCE: screen-printed carbon electrode; ISE: Ion-selective-electrodes; SVM-DA: support vector chamiche discriminant analysis; LDA: linear discriminant analysis;EEM: excitation-emission matrix; NMR: Nuclear magnetic resonance; PARAFA: parallel factor analysis; kNN: k-Nearest neighbours: PCA: principal component analysis; PLSDA: partial least squares discriminant analysis; OPS: ordered predictors selection; HS-SPME: headspace solid phase micro extraction; ANN-MLP: artificial neural network with multilayer perceptrons.