<![CDATA[The increasing demand for renewable energy has encouraged the optimization of palm oil biodiesel production to improve product quality and process efficiency. Biodiesel production is strongly influenced by operational parameters, including Free Fatty Acid (FFA) content, moisture content, methanol-to-oil molar ratio, catalyst concentration, reaction temperature, and reaction time, which may lead to quality variability and off-spec products. This study aimed to develop a predictive analysis model for palm oil biodiesel production using the Gaussian Naive Bayes algorithm based on a data mining approach. The study employed the Knowledge Discovery in Databases (KDD) framework using a secondary dataset consisting of 250 observations and six operational variables. Data preprocessing included missing value handling, Min-Max normalization, and Random Over-Sampling (ROS) to address class imbalance. The results showed that the model achieved an accuracy of 86.0%, weighted F1-score of 0.86, and cross-validation accuracy of 86.3 ± 2.4%. The analysis identified FFA, reaction temperature, and moisture content as the main factors influencing biodiesel quality. In addition, the model demonstrated high computational efficiency with a total processing time of 0.070 seconds, indicating its potential for real-time quality monitoring applications in biodiesel production systems.]]>
Copyrights © 2026
