Partial least squares and artificial neural networks for
multicomponent analysis from derivative UV-Vis spectra

D 2002

Partial least squares and artificial neural networks for multicomponent analysis from derivative UV-Vis spectra

TRNKOVÁ, Libuše, Eladia M. PEŇA-MENDEZ, Jana TOPINKOVÁ a Josef HAVEL

Základní údaje

Originální název

Partial least squares and artificial neural networks for multicomponent analysis from derivative UV-Vis spectra

Autoři

TRNKOVÁ, Libuše (203 Česká republika, garant), Eladia M. PEŇA-MENDEZ (724 Španělsko), Jana TOPINKOVÁ (203 Česká republika) a Josef HAVEL (203 Česká republika)

Vydání

1. vyd. Czech Republic, Brno, International Chemometric Conference - CHEMOMETRICS VI, od s. P21, 1 s. 2002

Nakladatel

Masaryk University, Brno, Czech Republic

Další údaje

Jazyk

angličtina

Typ výsledku

Stať ve sborníku

Obor

10406 Analytical chemistry

Stát vydavatele

Česká republika

Utajení

není předmětem státního či obchodního tajemství

Kód RIV

RIV/00216224:14310/02:00006479

Organizační jednotka

Přírodovědecká fakulta

ISBN

80-210-2918-8

Klíčová slova anglicky

Partial least squares (PLS);artificial neural networks(ANN);multicomponent analysis;derivative UV-Vis spectra;adenine;cytosine

Štítky

adenine, artificial neural networks(ANN), cytosine, derivative UV-Vis spectra, multicomponent analysis, Partial least squares (PLS)

Změněno: 30. 5. 2003 18:45, prof. RNDr. Libuše Trnková, CSc.

Anotace

V originále

This contribution presents a comparative study of the use of PLS and ANNs to analyze A and C mixtures using UV-Vis derivative spectra. The optimum ANN architecture enabling to model the system was established by means of TRAJAN 6.0 program. Several algorithms (Back propagation, Conjugate gradients, Quick propagation, and Delta-Bar Delta algorithm) were used for the training of the ANN to obtain a reliable model. With help of a suitable experimental design in combination with soft ANN modelling, the concentration of both A and C in mixtures can be quantified with an excellent accuracy (about 1 %). The quality of the testing set was evaluated on the basis of the average root mean square error for prediction (RMSEP) calculated from true and found values of A and C concentrations (RMSEP = 0.07 for A and 0.09 for C). It was found that ANN gives better results for the first and second derivative spectra than for original spectra. Furthermore, in comparison with PLS the ANN provides a more reliable and precise approach in the multicomponent analysis of A and C mixtures, where a number of different interactions take place.

Návaznosti

GA203/02/0422, projekt VaV

Název: Nové směry v elektrochemii nukleových kyselin a jejich aplikace v chemii životního prostředí

Investor: Grantová agentura ČR, Nové směry v elektrochemii nukleových kyselin a jejich aplikace v chemii životního prostředí

IAA1163201, projekt VaV

Název: Využití adsorptivní přenosové a eliminační techniky pro elektrochemickou analýzu oligonukleotidů a nukleových kyselin

Investor: Akademie věd ČR, Využití adsorptivní přenosové a eliminační techniky pro elektrochemickou analýzu oligonukleotidů a nukleových kyselin

MSM 143100011, záměr

Název: Struktura a vazebné poměry, vlastnosti a analýza syntetických a přírodních molekulových ansamblů

Investor: Ministerstvo školství, mládeže a tělovýchovy ČR, Struktura a vazebné poměry, vlastnosti a analýza syntetických a přírodních molekulových ansamblů

Citovat

TRNKOVÁ, Libuše, Eladia M. PEŇA-MENDEZ, Jana TOPINKOVÁ a Josef HAVEL. Partial least squares and artificial neural networks for multicomponent analysis from derivative UV-Vis spectra. In International Chemometric Conference - CHEMOMETRICS VI. 1. vyd. Czech Republic, Brno: Masaryk University, Brno, Czech Republic, 2002, s. P21, 1 s. ISBN 80-210-2918-8.

@inproceedings{405625,
   author = {Trnková, Libuše and PeňaandMendez, Eladia M. and Topinková, Jana and Havel, Josef},
   address = {Czech Republic, Brno},
   booktitle = {International Chemometric Conference - CHEMOMETRICS VI},
   edition = {1},
   keywords = {Partial least squares (PLS);artificial neural networks(ANN);multicomponent analysis;derivative UV-Vis spectra;adenine;cytosine},
   language = {eng},
   location = {Czech Republic, Brno},
   isbn = {80-210-2918-8},
   pages = {P21},
   publisher = {Masaryk University, Brno, Czech Republic},
   title = {Partial least squares and artificial neural networks for multicomponent analysis from derivative UV-Vis spectra},
   year = {2002}
}

TY  - JOUR
ID  - 405625
AU  - Trnková, Libuše - Peňa-Mendez, Eladia M. - Topinková, Jana - Havel, Josef
PY  - 2002
TI  - Partial least squares and artificial neural networks for multicomponent analysis from derivative UV-Vis spectra
PB  - Masaryk University, Brno, Czech Republic
CY  - Czech Republic, Brno
SN  - 8021029188
KW  - Partial least squares (PLS);artificial neural networks(ANN);multicomponent analysis;derivative UV-Vis spectra;adenine;cytosine
N2  - This contribution presents a comparative study of the use of PLS and ANNs to analyze A and C mixtures using UV-Vis derivative spectra. The optimum ANN architecture enabling to model the system was established by means of TRAJAN 6.0 program. Several algorithms (Back propagation, Conjugate gradients, Quick propagation, and Delta-Bar Delta algorithm) were used for the training of the ANN to obtain a reliable model. With help of a suitable experimental design in combination with soft ANN modelling, the concentration of both A and C in mixtures can be quantified with an excellent accuracy (about 1 %). The quality of the testing set was evaluated on the basis of the average root mean square error for prediction (RMSEP) calculated from true and found values of A and C concentrations (RMSEP = 0.07 for A and 0.09 for C). It was found that ANN gives better results for the first and second derivative spectra than for original spectra. Furthermore, in comparison with PLS the ANN provides a more reliable and precise approach in the multicomponent analysis of A and C mixtures, where a number of different interactions take place.
ER  -

TRNKOVÁ, Libuše, Eladia M. PEŇA-MENDEZ, Jana TOPINKOVÁ a Josef HAVEL. Partial least squares and artificial neural networks for multicomponent analysis from derivative UV-Vis spectra. In \textit{International Chemometric Conference - CHEMOMETRICS VI}. 1. vyd. Czech Republic, Brno: Masaryk University, Brno, Czech Republic, 2002, s.~P21, 1 s. ISBN~80-210-2918-8.

Podrobný výpis o publikaci