Comparison of Machine Learning Algorithms for Palm Oil Fresh Fruit Bunch (FFB) Ripeness

Determining the maturity level of FFB is crucial because it directly affects the quality and quantity of palm oil produced. Ripe FFB has high oil content, ensuring better-quality palm oil. Traditional methods include visual inspection, manual sampling, and physical testing, which are labor-intensive and subjective and can result in inconsistencies and errors. Machine learning algorithms can analyze datasets quickly and accurately, while also identifying patterns and features that are not easily visible to humans. Therefore, the aim of this study is to examine and evaluate the effectiveness of machine learning algorithm classifiers in determining oil palm FFB ripeness. The algorithms used in this research for classification analysis are Logistic Regression, Support Vector Machines (SVM), k-Nearest Neighbors (k-NN), Neural Networks (NN), and Naïve Bayes. In this research, the analysis was carried out using the Orange data mining tool, which carried out data analysis and data visualization. The results of performance evaluation was tested, and assessment (cross-validation accuracy estimation), ROC (Receiver Operating Characteristic) analysis, and confusion matrix. The best models were Neural Network, Logistic Regression and SVM. Naïve Bayes appears to have lower performance. The results using the prediction widget show that logistic regressing has the best accuracy.