<![CDATA[Pomelo (Citrus maxima), one of Indonesia's native citrus fruits, possesses high economic value and is widely cultivated across various regions in diverse varieties such as Bali Merah, Cikoneng, Nambangan, Raja, Ratu, and Pangkep. Despite its potential, quality assessment of pomelo fruits is still mostly conducted manually based on physical characteristics, which may lead to subjective and inconsistent results. This study aims to develop a more objective and efficient method by utilizing the K-Nearest Neighbor (K-NN) classification algorithm within a data mining framework. Six key features were used as classification variables: peel pigmentation, surface smoothness, fruit softness, weight, skin thickness, and overall quality. The research used a dataset collected from the North Sumatra Plantation Office over the past five years (2020–2024), which was processed and analyzed using the Orange application. Evaluation of the classification model achieved promising results, with an accuracy of 86.0%, F1-score of 0.860, precision of 0.861, recall of 0.860, AUC of 0.834, and MCC of 0.720. Additionally, predictions on new data samples confirmed the model’s ability to classify high-quality pomelo fruits effectively. These findings highlight the effectiveness of K-NN as a decision-support tool for improving fruit quality assessment processes and support the integration of data mining in smart agriculture practices.]]>
Copyrights © 2025
