Electronic tongues are devices used in the analysis of aqueous matrices for classification or quantification tasks. Sensor arrays in electronic tongues are composed of several sensors of different materials, a data acquisition unit and a pattern recognition system. These sensors can be of electrochemical type and one of the analytical methods carried out in the aqueous matrix corresponds to cyclic voltammetry. After performing the cyclic voltammetry, each sensor yields a voltammogram that relates the response in current to the change in voltage applied to the working electrode. A great amount of data is obtained in the experimental procedure. A conventional way to extract features of raw signals of the voltammograms obtained in an electronic tongue is by selecting specific features, related to the peaks recorded for the reduction and oxidation processes, such as potentials for the maximum current, maximum current values and widths of the observed peaks. In this work, a novel data processing methodology is developed for sensor arrays of a cyclic voltammetry electronic tongue. This methodology is composed of several stages such as data normalization through group scaling method and a nonlinear feature extraction step with locally linear embedding (LLE) technique. A reduced size feature vector is obtained that serves as input to a K-Nearest Neighbors (KNN) supervised classifier algorithm. A Leave-one-out cross validation procedure is performed to obtain the final classification accuracy. The methodology is validated with a data set of five different juices as liquid substances. Finally, 80% of classification accuracy was obtained.
- 88 Reads