The investigation was performed to test the potentials of the fingerprint clustering algorithm for a set of 1599 red wines in relation to some wine properties, comprised in the notion “wine quality”. We have obtained a distribution of the wines into different clusters as a result. Each cluster was composed of wine-objects with similar values of laboratory parameters and with a wine quality certificate. A correlation between the. quality of wines (a sensory taste factor) and the phisicochemical descriptors (laboratory analytical test results data) was observed and analyzed.
. Ebeler, S., Flavor Chemistry - Therty Years of Progress, Kluwer Academic Publishers, 1999, p. 409-422.
. A. Legin, A.; Rudnitskaya, L.; Luvova, Y.; Vlasov C., Natale and D’Amico, A., “Evaluation of Italian wine by the electronic tongue: recognition, quantitative analysis and correlation with human sensory perception”, Analytica Chimica Acta, 2003, 484(1), p. 33-34.
. Cortez, P.; Cerdeira, A.; Almeida, F.; Matos, T.; Reis, J., “Modeling wine preferences by data mining from hysicochemical properties”, http://repositorium.sdum.uminho.pt/bitstream/1822/10029/1/wine5.pdf
. Butina, D., “Unsupervised data base clustering based on Daylight’s fingerprint and Tanimoto similarity: A fast and automated way to cluster small and large data sets, J. Chem. Inf. Comput. Scie., 1999. 39, pp 747-750.