Bionic Electronic Nose Based on MOS Sensors Array and Machine Learning Algorithms Used for Wine Properties Detection

Abstract

In this study, a portable electronic nose (E-nose) prototype is developed using metal oxide semiconductor (MOS) sensors to detect odors of different wines. Odor detection facilitates the distinction of wines with different properties, including areas of production, vintage years, fermentation processes, and varietals. Four popular machine learning algorithms—extreme gradient boosting (XGBoost), random forest (RF), support vector machine (SVM), and backpropagation neural network (BPNN)—were used to build identification models for different classification tasks. Experimental results show that BPNN achieved the best performance, with accuracies of 94% and 92.5% in identifying production areas and varietals, respectively; and SVM achieved the best performance in identifying vintages and fermentation processes, with accuracies of 67.3% and 60.5%, respectively. Results demonstrate the effectiveness of the developed E-nose, which could be used to distinguish different wines based on their properties following selection of an optimal algorithm.

Air Quality Index and Air Pollutant Concentration Prediction Based on Machine Learning Algorithms

“…where M denotes the total number of features [32]. RF can be applied to classification and regression problems, depending on whether the trees are classification or regression trees.…”

Beverage and Food Fragrance Biotechnology, Novel Applications, Sensory and Sensor Techniques: An Overview

“…E-nose devices can be classified according the origin of their signal-transduction mechanisms. Several signal-transduction mechanisms have been reported: Metal oxide semiconductor (MOS) [148,149]; conducting polymer films [150]; acoustic wave devices [151]; optical transducers [152,153]; electrochemical systems [154,155] polymer film chemo-resistors [156] and quartz microbalance (QMB) sensors [157]. Regardless of the transduction mechanism, the higher the number of sensory elements in a cross-reactive sensor array (CRSA), the better is the data collected, and the specificity of the analyte identification and classification.…”

“…The deviations normally caused by environmental and physicochemical influences disrupt the compatibility between the gas-sensor responses and the artificial intelligence algorithms [159]. Drift component can be decomposed based on statistical characteristics by means of multivariate statistical analysis such as principal component analysis (PCA) and PCA-based component corrections [160,161] independent component analysis [162], wavelet [148] and, more recently, by active learning on dynamic clustering [159] Besides e-noses, science has arranged the manner of also mimic the Human tongue. The first e-tongue appeared in the 1990s.…”

Assessment of Smoke Contamination in Grapevine Berries and Taint in Wines Due to Bushfires Using a Low-Cost E-Nose and an Artificial Intelligence Approach

“…Ayhan et al [ 12 ] explored the fluctuation-enhanced sensing method to detect and classify gases with improved accuracy when developing classification models using machine learning algorithms. Some applications include medical diagnostics [ 13 ], space shuttles and stations [ 14 , 15 , 16 ], crime and security [ 17 ], and food and beverages, such as rapeseed to detect volatile compounds in pressed oil [ 18 ], wine [ 19 ], and beer [ 20 ], among others. The latter study describes a low-cost e-nose developed with nine gas sensors to assess the aroma profile of beers coupled with machine learning modeling.…”