Quality Classification of Palm Oil Varieties Using Naive Bayes Classifier

Abstract

As one of the leading commodities of the Indonesian economy, the ever-increasing production of palm oil has created intense competition among palm oil (CPO) producers. This causes CPO producers to increase their palm oil production without compromising the quality of the palm oil produced. CPO producers are required to be able to objectively determine the quality of superior and precise oil palm varieties in order to produce high economic value palm oil. Therefore, a model is needed to determine the quality of oil palm from several existing varieties. The Naive Bayes Classifier method in this study was used to classify the quality of oil palm based on predetermined variables using a data set of 28 oil palm varieties. Method testing is done by using a confusion matrix and K-fold cross-validation scheme. This study shows a reasonably high accuracy value of 64.25% and a low error rate of 35.7%, indicating that the Naive Bayes Classifier can classify the quality of oil palm varieties quite well.