<![CDATA[The study focuses on an examination of the applicability of the Support Vector Machine (SVM) algorithm and its implication for the classification of the quality of the Crude Palm Oil (CPO) produced by PT. PP London Sumatra Indonesia Tbk. The authentic quality parameters: Water (VM), Dirt, and Free Fatty Acid (FFA) were chosen to train the SVM model which was tested on the data of 2020–2022 and containing 1,095 records. The research utilized Google Colab Python Notebooks for the analysis of data, resulting to an accuracy of 84. 15%. This indicates that SVM is a reliable technique to work with complicated, multi variet data,; which can be quite helpful in the CPO quality classification, where traditional algorithms may not be efficient. Data preprocessing including normalization and outlier detection has been cited as some of the ways that would improve the performance of the model as highlighted in this study. Comparing the results with other machine learning algorithms such as Random Forest and Neural networks proved the efficiency of SVM even though there were misclassification made. The result also suggests that SVM has a strong capability to support the quality assurance activities in the palm oil industry by eliminating human intervention and increasing the working productivity. Further study could continue in the directions of incorporating the SVM model with other methods of machine learning for even better enhancement of the CPO quality assessment.]]>
Copyrights © 2024
