Authentication of Galician (NW Spain) quality brand potatoes using metal analysis. Classical pattern recognition techniques versus a new vector quantization-based classification procedure

Citation
Rm. Pena et al., Authentication of Galician (NW Spain) quality brand potatoes using metal analysis. Classical pattern recognition techniques versus a new vector quantization-based classification procedure, ANALYST, 126(12), 2001, pp. 2186-2193
Citations number
35
Categorie Soggetti
Chemistry & Analysis","Spectroscopy /Instrumentation/Analytical Sciences
Journal title
ANALYST
ISSN journal
00032654 → ACNP
Volume
126
Issue
12
Year of publication
2001
Pages
2186 - 2193
Database
ISI
SICI code
0003-2654(2001)126:12<2186:AOG(SQ>2.0.ZU;2-J
Abstract
The objective of this work was to develop a classification system in order to confirm the authenticity of Galician potatoes with a Certified Brand of Origin and Quality (CBOQ) and to differentiate them from other potatoes tha t did not have this quality brand. Elemental analysis (K, Na, Rb, Li, Zn, F e, Mn, Cu, Mg and Ca) of potatoes was performed by atomic spectroscopy in 3 07 samples belonging to two categories, CBOQ and Non-CBOQ potatoes. The 307 x 10 data set was evaluated employing multivariate chemometric techniques, such as cluster analysis and principal component analysis in order to perf orm a preliminary study of the data structure. Different classification sys tems for the two categories on the basis of the chemical data were obtained applying several commonly supervised pattern recognition procedures [such as linear discriminant analysis, K-nearest neighbours (KNN), soft independe nt modelling of class analogy and multilayer feed-forward neural networks]. In spite of the fact that some of these classification methods produced sa tisfactory results, the particular data distribution in the 10-dimensional space led to the proposal of a new vector quantization-based classification procedure (VQBCP). The results achieved with this new approach (percentage s of recognition and prediction abilities > 97%) were better than those att ained by KNN and can be compared advantageously with those provided by LDA (linear discriminant analysis), SIMCA (soft independent modelling of class analogy) and MLF-ANN (multilayer feed-forward neural networks). The new VQB CP demonstrated good performance by carrying out adequate classifications i n a data set in which the classes are subgrouped. The metal profiles of pot atoes provided sufficient information to enable classification criteria to be developed for classifying samples on the basis of their origin and brand .