Improving the geographical traceability of tea in China based on stable isotope ratios

The potential of improving the classification of tea samples from different regions was studied by using stable isotope ratios analysis. The stable isotope ratios of 44 elements in tea samples were determined (p < 0.05).The results showed that 34 stable isotopes ratios were statistically significant, and tea in the four regions had their own characteristic variables. PCA, HCA, PLS-DA, BP-ANN and LDA were used to analyze the stable isotope ratio data in tea. Six key variables were identified to provide the greatest difference between the samples. The overall correct classification rate, cross validation rate and blind validation rate given by LDA are all 100