<![CDATA[Universities are responsible for producing quality graduates and reducing dropout rates (DO), a serious challenge for the Widuri Institute of Social and Technology (ISTEK). This phenomenon has a negative impact on the quality of education and accreditation, making early identification of students who have the potential to drop out (DO) very crucial. This study aims to apply the Naïve Bayes algorithm to predict the potential for dropout (DO) of ISTEK Widuri students based on data on the activities of the 2021, 2022, and 2023 intakes. Naïve Bayes has proven effective in classifying students at risk of dropping out (DO). The Semester Credit Unit (SKS) attribute is the most dominant indicator, students with low SKS have a high potential for dropping out (DO). Model performance varies for each batch, in the 2021 batch it reached 90% accuracy (100% DO precision, 40% recall), the 2022 batch showed 93.75% accuracy (100% DO precision, 60% DO recall), and the 2023 batch had 86.67% accuracy (100% DO precision, 33.33% DO recall). This model is very good at validating students who are safe from DO (100% recall of Not DO in all batches). Even so, the model still needs to be improved so that it can find all students who are at risk of dropping out (DO) as a whole. The prediction results for students with the potential for DO at ISTEK Widuri Jakarta are expected to support more optimal prevention efforts and contribute to improving the quality of education.]]>
Copyrights © 2025
