Oil Palm Fresh Fruit Bunches Maturity Prediction by Using Optical Spectrometer. Issue 1 (July 2020)

• Record Type: Journal Article
• Authors: Tuerxun, Adilijiang; Mohamed Shariff, Abdul Rashid; Janius, Rimfiel; Abbas, Zulkifly; Mahdiraji, Ghafour Amouzad
• Is Part Of: IOP conference series. Volume 540:Issue 1(2020)
• Publication Date: 2020-07
• Subjects: Data Mining -- Maturity Classification -- Oil Palm -- Spectroscopy; Earth sciences -- Periodicals; Environmental sciences -- Congresses; Environmental sciences -- Periodicals; 550.5
• DOI: 10.1088/1755-1315/540/1/012085
• Languages: English
• ISSNs: 1755-1307

Abstract

Abstract: The accuracy in grading the maturity level of oil palm fruits is one of the important factors for the milling industry since it can directly affect the oil palm yield. Numerous studies have investigated for oil palm maturity classification using different approaches. However, there is a lack of clear guidance on which algorithm or technique to be applied for the oil palm maturity classification under different situations. In this research, several machine learning algorithms in the Weka data mining tool were applied for building classifier models and tested on samples of fresh fruit bunch (FFB). The maturity grade determined from the extracted sample features under the spectrometer is compared to the results of the manual standard grader of the milling industry. The validated result shows that simple lazy KStar algorithm performs better with 63% classifier model performance and the weighted average of receiver operating characteristics (ROC) curve area of 83.2%, which established the classifier model accuracy and enhances the model generalization when applied to the new set of test data for validation. The results of these studies make a contribution to the ongoing research in evaluating new techniques for oil palm maturity determination.