University of Tehran Press Iranian Journal of Biosystem Engineering 2008-4803 50 3 2019 10 23 Classification of Different Iranian Rice Varieties and Frauded Rice Based on Volatile Compounds Detected by Electronic Nose Method Classification of Different Iranian Rice Varieties and Frauded Rice Based on Volatile Compounds Detected by Electronic Nose Method 595 606 73653 10.22059/ijbse.2019.263221.665081 FA Bahlool Ramesh M.Sc. Graduated, Faculty of Agricultural Engineering and Technology, University of Tehran, Karaj, Iran Seyed Saeid Mohtasebi Professor, Faculty of Agricultural Engineering and Technology, University of Tehran, Karaj, Iran 0000-0002-4031-1095 Shahin Rafiee Professor, Faculty of Agricultural Engineering and Technology, University of Tehran, Karaj, Iran Journal Article 2018 08 16 Rice aroma is one of the important features of rice quality which affects its marketability. In this study, an electronic system consisting of six semiconductor metal oxide sensors was used as a non-destructive method for the separation of Iranian rice varieties and a frauded rice sample, which is a kind of common fraud in rice supply. Analysis of PCA with two main components covered 89% of the variance (variation) of the data for five original rice samples. Also, described 96% of the variance of data for four rice samples, which included two varieties of rice and two fraud samples using LDA method with the accuracy of 100%. The precision of the ANN method was obtained as 98.6% for separation of the two groups of Iranian varieties and the frauded samples. Rice aroma is one of the important features of rice quality which affects its marketability. In this study, an electronic system consisting of six semiconductor metal oxide sensors was used as a non-destructive method for the separation of Iranian rice varieties and a frauded rice sample, which is a kind of common fraud in rice supply. Analysis of PCA with two main components covered 89% of the variance (variation) of the data for five original rice samples. Also, described 96% of the variance of data for four rice samples, which included two varieties of rice and two fraud samples using LDA method with the accuracy of 100%. The precision of the ANN method was obtained as 98.6% for separation of the two groups of Iranian varieties and the frauded samples. https://ijbse.ut.ac.ir/article_73653_c8c01ee70f024b956b338656f7261344.pdf