Application of the Naive Bayes Method for Determining the Quality of Crude Palm Oil (CPO) at PTPN 2 Sawit Seberang

Abstract

The palm oil industry is a vital pillar of Indonesia's economy, with Crude Palm Oil (CPO) as one of its leading commodities. The quality of CPO significantly impacts its competitiveness and market price internationally. PTPN 2 Sawit Seberang, as a prominent CPO processing company, faces challenges in consistently maintaining product quality. Key factors affecting CPO quality include moisture content, free fatty acids, and impurity levels, which are difficult to manage manually. To address these challenges, this study applies the Naive Bayes method as an efficient and fast classification tool for determining CPO quality. Naive Bayes was chosen for its simplicity in probability calculations and its ability to handle data classification with reasonable accuracy. The data used in this study include moisture content, free fatty acids, and impurity levels measured between February and June 2024. The data was split into training data (80%) and testing data (20%) and analyzed using RapidMiner software. The results show that the Naive Bayes method achieved an accuracy rate of 66.6%, with precision and recall values of 50% each. Although the accuracy could be improved, the application of this method has significantly enhanced the efficiency of determining CPO quality. Thus, the implementation of the Naive Bayes method in determining CPO quality at PTPN 2 Sawit Seberang is an effective step towards improving operational efficiency, classification accuracy, and decision-making quality related to product standards, ultimately supporting the company's competitiveness in the global market.