<![CDATA[Higher education institutions generally have not optimally utilized student academic data as an early detection tool, so students at risk of declining academic performance are often identified too late. This study addresses this problem by applying the Decision Tree algorithm, based on entropy and information gain calculations, to classify student academic performance into three categories (Good, Sufficient, and Poor) using Grade Point per Semester (IPS1–IPS8) and Cumulative Grade Point Average (IPK) features from 518 student records. The purpose of this study is to build a classification model that is both accurate and interpretable, so that its contribution can be used by universities as a transparent decision-support tool for early academic intervention, unlike black-box models that are difficult to explain. The dataset was divided using a 70:30 training-testing ratio. Preliminary results show that the IPK attribute is the most influential factor, with the highest Information Gain of 0.8766, selected as the root node, while model evaluation on the testing data using a confusion matrix yields an accuracy of 96.15%, precision 94.71%, recall 95.10%, and F1-score 94.87%. These results indicate that the constructed Decision Tree model is suitable for use as an early-detection tool for students requiring academic attention.]]>
Copyrights © 2026
