Application of Cluster Analysis in Raw Material Formula of Paper-making Reconstituted Tobacco

24 raw material samples of paper-making reconstituted tobacco were selected and analyzed by SDE-GC/MS, and the data were classified using cluster analysis by DPS software. The result shows that the samples could be classified as two different categories. Tobacco stem and tobacco scrap belonged to different categories. Samples from same region were more similar. After the smoke panel test, the sensory result was in accord with the analytic result.