Classification of chemical analytical data with the use of a combination of the Kohonen and the probabilistic neural networks
Abstract
The paper presents a novel procedure capable to classify objects proceeding from their chemical characteristics. The approach is based on a combinantion of the Kohonen and the probabilistic neural networks. In contrast to existing analogs, the procedure does not require any a priori information about the number of classes and the patterns in the training set. To verify the procedure, the problem of the classification of water samples from different Kharkiv springs and rivers has been considered. The initial experimental data set consisted of concentrations of metal ions in water samples.
Downloads
References
Vlasov Yu., Legin A., Rudnitskaya A., Di Natale C., D’Amico A. Nonspecific sensor arrays (“electronic tongue”) for chemical analysis of liquids // Pure Appl. Chem. 2005. 77(11). P. 1965.
Hardcastle W. A. Qualitative analysis: a guide to best practice. Cambridge: Royal Society of Chemistry, 1998, 24 p. ISBN 0-85404-462-0.
Milman B. L. Identification of chemical compounds // Trends Anal. Chem. 2005. 24(6). P. 493.
Rodionova O. Ye., Pomerantsev A. L. Hemometrika v analiticheskoy himii, 2006, 61 p. http://www.chemometrics.ru/materials/articles/chemometrics_review.pdf [in Russian]
Adams M. J. Chemometrics in analytical spectroscopy (2nd ed.). Cambridge: Royal Society of Chemistry, 2004, 238 p. ISBN 0-85404-555-4.
Mutihac L., Mutihac R. Mining in chemometrics // Anal. Chim. Acta. 2008. 612. P. 1.
de Juan A., Fonrodona G., Casassas E. Solvent classification based on solvatochromic pa-rameters: a comparison with the Snyder approach // Trends Anal. Chem. – 1997. 16(1). P. 52.
Simeonov V., Simeonova P., Tsakovskii S., Lovchinov V. Lake water monitoring data as-sessment by multivariate statistics // J. Water Resource Protect. 2010. 2. P. 353.
Skorek R., Jablonska M., Polowniak M., Kita A., Janoska P., Buhl F. Application of ICP-MS and various computational methods for drinking water quality assessment from the Silesian District (Southern Poland) // Centr. Eur. J. Chem. 2010. 10(1). P. 71.
Balabin R. M., Safieva R. Z., Lomakina E. I. Gasoline classification using near infrared (NIR) spectroscopy data: comparison of multivariate techniques // Anal. Chim. Acta. 2010. 671. P. 27.
Galao O. F., Borsato D., Pinto J. P., Visentainer J. V., Carrao-Panizzi M. C. Artificial neural networks in the classification and identification of soybean cultivars by planting region // J. Braz. Chem. Soc. 2011. 22(1). P. 142.
Pushkarova Ya., Kholin Yu. The classification of solvents based on solvatochromic character-istics: the choice of optimal parameters for artificial neural networks // Centr. Eur. J. Chem. 2012. 10(4). P. 1318.
Krasnianchyn Ya. N., Panteleimonov A. V., Kholin Yu. V. // Visn. Hark. nac. univ., 2010, № 895, Ser. Him., issue. 18(41), P. 39. [ISSN 2220-637X (print), ISSN 2220-6396 (online), http://chembull.univer.kharkov.ua/archiv/2010/05.pdf] [in Russian].
Krasnianchyn Ya. N., Panteleimonov A. V., Kholin Yu. V. // Visn. Hark. nac. univ., 2010, № 932, Ser. Him., issue. 19(42), P. 170. [ISSN 2220-637X (print), ISSN 2220-6396 (online), http://chembull.univer.kharkov.ua/archiv/2010_2/21.pdf] [in Russian].
Kruglov V. V., Borisov V. V. Iskusstvenny'e neyronny'e seti. Teoriya i praktika. M.: Gorya-chaya liniya-Telekom, 2002, 382 p. ISBN 5-93517-031-0. [in Russian]
Osowski S. Sieci neuronowe do przetwarzania informacji. Warszawa: Oficyna wydawnicza politechniki Warszawskiej, 2000. ISBN 83-7207-187-X.
Dong M., Wang N. Adaptive network-based fuzzy inference system with leave-one-out cross-validation approach for prediction of surface roughness // Appl. Math. Model. 2011. 35(3). P. 1024.
Pushkarova Ya. N., Sledzevska A. B., Panteleimonov A. V., Titova N. P., Yurchenko O. I., Ivanov V. V., Kholin Yu. V. // Moscow Univ. Chem. Bull. 2012. 67(6). P. 287.
Sharaf M. A., Illman D. L., Kowalski B. R. Chemometrics. New York, Chichester, Brasbane, Totonto, Singapore: John Wiley & Sons, 1986, 332 p. ISBN 0471831069.
