Compression of electronic tongue data based on voltammetry — a comparative study

摘要: Abstract In this paper, three data compression methods are investigated to determine their ability reduce large sets obtained by a voltammetric electronic tongue without loss of information, since compressed will save storage and computational time. The is based on combination non-specific sensors pattern recognition tools, such as principal component analysis (PCA). A series potential pulses decreasing amplitude applied one working electrode at time resulting current transients collected each step. Voltammograms containing up 8000 variables subsequently obtained. wavelet transformation (WT) hierarchical (HPCA). Also, new chemical/physical model theory developed in order extract interesting features the transients, revealing different information about species solutions. Two experiments performed, solutions electroactive compounds other complex samples, juices from fruits tomatoes. It shown that WT HPCA compress improves separations slightly. able two largest extent, 16 variables. When scaled unit variance, separation even further for model.