<![CDATA[This study aims to identify the authentication of civet coffee using a Soft independent modeling of class analogy(SIMCA) method and principal component analysis (PCA). The test carried out on the coffee powder measuring0.297 millimeters (mesh 50). Comparison of blend that is samples 1- 50 each 1 g of pure civet coffee, samples51- 60 each 0.9 g civet coffee and 0.1 g arabica coffee, samples 61-70 each 0.8 g civet coffee and 0.2 g arabicacoffee, samples 71-80 each 0.7 g civet coffee and 0.3 g arabica coffee, samples 81-90 each 0.6 g civet coffee and0.4 g arabica coffee, samples 90-100 each 0.5 g civet coffee and 0.5 g arabica coffee. The classification resultsshow SIMCA and PCA methods are able to identify civet coffee mixture. PC 1 explains 75% the variance of dataand PC2 explains 17% the variance of data. Values obtained on SIMCA classification are specificity 76%,sensitivity of 84% and accuracy of 80%, with a value error of 23%.Keywords: Arabica coffee,civet coffee, PCA, SIMCA, UV-Vis spectroscopy.]]>
Copyrights © 2016
