Applying Machine Learning to High-Quality Wine Identification
This paper discusses a machine learning approach, aimed at the definition of methods for authenticity assessment of some of the highest valued Nebbiolo-based wines from Piedmont (Italy). This issue is one of the most relevant in the wine market, where commercial frauds related to such a kind of products are estimated to be worth millions of Euros. The main objective of the work is to demonstrate the effectiveness of classification algorithms in exploiting simple features about the chemical profile of a wine, obtained from inexpensive standard bio-chemical analyses. We report on experiments performed with datasets of real samples and with synthetic datasets which have been artificially generated from real data through the learning of a Bayesian network generative model.
KeywordsClassification Fraud detection Artificial data generation
The present work has been funded by Regione Piemonte (POR-FESR grants), as a part of the TRAQUASwine project. We would like to thank M. Arlorio, J.D. Coïsson, M. Locatelli and F. Travaglia for their collaboration.
- 1.Arlorio, M., Coisson, J., Leonardi, G., Locatelli, M., Portinale, L.: Exploiting data mining for authenticity assessment and protection of high-quality Italian wines from Piedmont. In: Proceedings of 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 2015), pp. 1671–1680. AUS, Sydney (2015)Google Scholar
- 6.Grzegorczyk, M.: An introduction to Gaussian Bayesian networks. In: Yan, Q. (ed.) Systems Biology in Drug Discovery and Development, vol. 662, pp. 121–147. Springer, Heidelberg (2010). https://doi.org/10.1007/978-1-60761-800-3_6
- 9.Hall, M.A.: Correlation-based feature subset selection for machine learning. Ph.D. thesis, University of Waikato, Hamilton, New Zealand (1998)Google Scholar
- 11.Locatelli, M., Travaglia, F., Coïsson, J., Bordiga, M., Arlorio, M.: Phenolic composition of Nebbiolo grape (Vitis vinifera L.) from Piedmont: characterization during ripening of grapes selected in different geographic areas and comparison with Uva Rara and Vespolina. Eur. Food Res. Technol. 242, 1057–1068 (2016)CrossRefGoogle Scholar
- 13.Mattera, D., Haykin, S.: Support vector machines for dynamic reconstruction of a chaotic system. In: Schölkopf, B., Burges, C., Smola, A. (eds.) Advances in Kernel Methods, pp. 211–241. MIT Press, Cambridge (1999)Google Scholar
- 14.Platt, J.: Fast training of support vector machines using sequential minimal optimization. In: Schölkopf, B., Burges, C., Smola, A. (eds.) Advances in Kernel Methods, pp. 185–208. MIT Press, Cambridge (1999)Google Scholar
- 15.Platt, J.: Probability for SV machines. In: Smola, A., Batlett, P., Schölkopf, B., Schuurmans, D. (eds.) Advances in Large Margin Classifiers, pp. 61–74. MIT Press, Cambridge (2000)Google Scholar
- 16.Spirtes, P., Glymour, C., Scheines, R.: Causation, Prediction and Search. Springer, Berlin (1993). https://doi.org/10.1007/978-1-4612-2748-9
- 19.Wagstaff, K.: Machine learning that matters. In: Proceedings of the 29th International Conference on Machine Learning (ICML 2012), Edinburgh, UK (2012)Google Scholar