<![CDATA[Tobacco plants are plantation products but not food crops, their leaves are usually used as the main ingredient in the making of cigarettes and cigars. Tobacco cultivation has been known for a long time in Indonesia, the cultivation of tobacco from generation to generation has resulted in the emergence of many new varieties in various regions in Indonesia. The number of tobacco varieties can be grouped by cultivation and type. The large number of tobacco varieties makes it difficult for farmers to distinguish the types of tobacco plants because the morphology and biology between tobacco plants are almost similar, so to make it easier to determine the type of tobacco plants, a system with a classification method is needed. One of the classification methods is the Naive Bayes algorithm. In this study, 11 classes were used and 19 features were used. In addition to classification, the feature selection method is also used to get a good combination of features and accuracy values, Information Gain used as the feature selection method. In the evaluation, the K-fold cross validation method is used to eliminate doubts on the data with k = 10. The result of all the tests carried out, the highest average accuracy for all balanced class tests was 52.72% using 17 features. Meanwhile, the highest accuracy of all unbalanced class tests is 64.06% when using 15 features.]]>
Copyrights © 2021
