Classification and authentication of unknown water samples using machine learning algorithms

Electronic tongue: a versatile tool for mineral and fruit-flavored waters recognition

“…So, monitoring the quality of these waters, during their production line as well as of the final bottled product is of utmost importance. However, the characterization of liquid substances is often inexact due to nonhomogeneity of the species in the liquid and to fluctuations rose due to the slight variations of the surface or ground raw water [10]. In 2007, Ciosek and Wróblewski [11] reviewed several applications of potentiometric and voltammetric electronic tongue (E-tongue) sensor arrays for liquid sensing, showing that E-tongue devices have been successfully used for water analysis in food, environmental and industrial fields (e.g., mineral waters recognition, assessment of ion levels and total hardness in natural waters, analysis of pollutants (metal ions, organic compounds or microorganisms) in natural waters, monitoring of spoilage of seawater and freshwater fish, analysis of paper-mill wastewaters, determination of anionic and nonionic surfactants in natural waters).…”

“…Indeed in the last decade, the use of potentiometric or voltammetric E-tongues has been increasingly reported since they have been proved to be useful in assessment of complex liquid media. Regarding drinking water analysis, E-tongues are mainly applied for: (i) monitoring water quality and detection of specific chemical compounds, possible contaminants, microorganisms, determination of sewage water concentrations in drinking water [1,7,[12][13][14][15][16][17][18]; or, (ii) water classification/authentication according to commercial brands or type of water [10,[19][20][21][22][23][24][25][26][27][28]. Water safety monitoring technologies based on E-tongues have been recently discussed by Vagin and Winquist [29], which demonstrated the potential of E-tongues for high throughput monitoring of edible or potable products.…”

“…A potentiometric E-tongue system with multi-electrodes assembly was also capable to recognize different classes of drinking water namely mineral, sparkling and tap water [19]. A voltammetric E-tongue with two working electrodes (Ag and Pt) was also reported for mineral water classification and authentication [10]. A novel E-tongue based on Fourier transform impedance spectroscopy was also used for classifying bottled mineral waters according to brand [24].…”

The e‐tongue‐based classification and authentication of mineral water samples using cross‐correlation‐based PCA and Sammon's nonlinear mapping

“…Multidimensional discriminant analysis, fuzzy clustering and SVM are computationally expensive techniques. Therefore, Kundu et al designed a PLS-based classifier for mineral water sample classification and authentication using e-tongue signatures [40]. PLS has also been applied to classification problems by encoding class memberships in an appropriate indicator matrix.…”

Discrimination of mineral waters by electronic tongue, sensory evaluation and chemical analysis

“…PCA showed that the discrimination was made according to the ionic content in water. Kundu, Panchariya, and Kundu (2011) introduced a development of water sample classification and authentication based on an electronic tongue application with pulse voltametry method. The results of the system, implemented in a laboratory environment for six different certified water samples, were used to build classification and regression models.…”