Wu, Hao,Tian, LChen, Bo,Jin, BTian, BinXie, Liqi,Rogers, KMLin, G2019-07-292019-07-062019-12-152019-07-050308-814631362191 (pubmed)https://hdl.handle.net/10182/10843Multi-isotope and multi-elemental analyses were performed on 600 red wine samples imported into China from 7 different countries and compared with Chinese wine. Carbon and oxygen isotopes and 16 elements were used to determine origin traceability. Our goal was to build a classification tool using data modeling that can verify the geographic origin of wines imported into China. Multivariate analyses of the isotopic and elemental data revealed that it is possible to determine the geographical origin for most imported wines with a high level of confidence (> 90%). The results show that Artificial Neural Network method had a high discrimination accuracy and is more suitable than Discrimination Analysis and Random Forest methods when it comes to classifying wine origin on a global scale. In conclusion, stable isotope and trace element analyses followed by multivariate processing of the data is a fast and efficient technique suitable for global wine traceability.8 pagesPrint-Electronicen© 2019 Elsevier Ltd. All rights reserved.trace elementsstable isotopesgrape winegeographical originfraud detectionTrace ElementsCarbon IsotopesOxygen IsotopesMultivariate AnalysisDiscriminant AnalysisModels, StatisticalFood AnalysisWineChinaMass SpectrometryJournal Article10.1016/j.foodchem.2019.125137ANZSRC::0908 Food SciencesANZSRC::090806 Wine Chemistry and Wine Sensory Science1873-7072ANZSRC::3006 Food sciences