Artificial neural networks for quantification in unresolved capillary electrophoresis peaks

Varování

Publikace nespadá pod Filozofickou fakultu, ale pod Přírodovědeckou fakultu. Oficiální stránka publikace je na webu muni.cz.
Autoři

BOCAZ-BENEVENTI Gaston LATORRE Rosa FARKOVÁ Marta HAVEL Josef

Rok publikování 2002
Druh Článek v odborném periodiku
Časopis / Zdroj Analytica Chimica Acta
Fakulta / Pracoviště MU

Přírodovědecká fakulta

Citace
Obor Analytická chemie, separace
Klíčová slova capillary zone electrophoresis; unresolved peaks; experimental design; normalization; artificial neural networks; quantitation
Popis The application of the combination of experimental design (ED) and artificial neural networks (ANNs) for the quantification of overlapped peaks in capillary zone electrophoresis is described. When the total separation cannot be achieved by separation techniques, the use of ED-ANN can be a suitable approach. The unstability of EOF causes peak shift that has to be corrected in order to apply ED-ANN methods. In this work, normalization procedure of electropherograms with consequent application of ANNs for quantification purpose was developed. Both, spectra and electropherograms can be used as multivariate data. In general, both kinds of data were found to be suitable for unresolved peaks quantification by ED-ANN approach.
Související projekty:

Používáte starou verzi internetového prohlížeče. Doporučujeme aktualizovat Váš prohlížeč na nejnovější verzi.