Pisco is an alcoholic beverage obtained from grape juice distillation. Considered the flagship drink of Peru, it is produced following strict and specific quality standards. In this work, sensing results for volatile compounds in pisco, obtained with an electronic nose, were analyzed through the application of machine learning algorithms for the differentiation of pisco varieties. This differentiation aids in verifying beverage quality, considering the parameters established in its Designation of Origin”. For signal processing, neural networks, multiclass support vector machines and random forest machine learning algorithms were implemented in MATLAB. In addition, data augmentation was performed using a proposed procedure based on interpolation–extrapolation. All algorithms trained with augmented data showed an increase in performance and more reliable predictions compared to those trained with raw data. From the comparison of these results, it was found that the best performance was achieved with neural networks.

There is great interest in the development of a simple system that could identify adulteration or counterfeiting of Peruvian Pisco (a grape-based alcoholic drink). In this study, sensors based on SnO₂-TiO₂ and SnO₂-MoO₃ composites with different weight composition ratios were synthesized and characterized. These sensors were tested with aqueous solutions of EtOH/MeOH and Pisco samples of Italia and Quebranta varieties in order to explore their capacity to identify variations in these beverages. The response profile of the most sensitive sensors showed an enhanced response to alcoholic samples with greater content of ethanol up to a concentration of 45%v/v, while the increased content of methanol in the range of 0.1 to 0.3% v/v diminished the intensity of the sensor response. Differences in the composition of methanol and ethanol in the Pisco varieties studied (Italia and Quebranta) were correlated to the capacity of the composite-based sensors to differentiate them with greater performance. Sensors based on SnO₂-TiO₂-1/2 composites showed greater reproducibility in their response profile over time in comparison to SnO₂-TiO₂-1/1 and SnO₂-MoO₃ composites. The PCA showed that composite sensors were able to differentiate Pisco samples according to the varieties, Italia and Quebranta.