Determination of the Wine Variety and Geographical Origin of White Wines Using Neural Network Technologies
- 18 Downloads
In order to determine the geographical origin and wine variety of white wines, we studied 153 samples of the white wines Riesling (49), Chardonnay (56), and Muscat (48) produced in the territory of the main wineries of geographical zones in the Krasnodar krai. The concentrations of trace and macro elements in wines were determined by inductively coupled plasma atomic emission spectrometry. Chemometric studies were performed using the STATISTICA Neural Networks. From a set of 15 trace and macro elements determined, 5 trace elements (Fe, Mg, Rb, Ti, and Na) were recognized by correlation analysis as the predictors of a constructed neural network model, which successfully identified the brands of wines. To determine the region of grape growing, a neural network model was constructed based on six predictors: five trace elements and a specified wine brand. A software was developed to automate the computations required.
Keywords:geographical origin and wine variety elemental analysis neural network technologies
The experiments were carried out with the use of equipment of the Environmental Analytical Center of Collective Use at the Kuban State University, unique identifier RFMEFI59317Х0008.
This study was supported by the Russian Foundation for Basic Research (project no. 18-03-00059).
- 10.Kaunova, A.A. Titarenko, V.O., Temerdashev, Z.A., Sekunova, M.V., and Popandopulo, V.G., Zavod. Lab., Diagn. Mater., 2016, vol. 82, no. 8, p. 69.Google Scholar
- 11.Giaccio, M. and Vicentini, A., J. Commod. Sci., Technol. Qual., 2008, vol. 47, p. 267.Google Scholar
- 16.Rapeanu, G., Vicol, C., and Bichescu, C., Innovative Rom. Food Biotechnol., 2009, vol. 5, p. 1.Google Scholar
- 20.Food Protected Designation of Origin: Methodologies and Applications, de la Guardia, M. and Gonzálvez, A., Eds., vol. 60 of Comprehensive Analytical Chemistry, Amsterdam: Elsevier, 2013.Google Scholar
- 23.Hill, T. and Lewicki, P., Statistics Methods and Applications, Tulsa, OK: StatSoft, 2007.Google Scholar
- 24.Neironnye seti. STATISTICA Neural Networks. Metodologiya i tekhnologii sovremennogo analiza dannykh (Neural Networks. STATISTICA Neural Networks. Methodology and Technology of Modern Data Analysis), Borovikov, V.P., Ed., Moscow: Telekom, 2008, 2nd ed.Google Scholar