Klasifikasi Kelayakan Bantuan Pendidikan Menggunakan Metode Decision Tree

Abstract

This study aims to implement the Decision Tree method to classify scholarship eligibility based on student data. The dataset used consisted of 4,424 student records with 35 numerical attributes covering academic, administrative, and socioeconomic aspects. The preprocessing stage included data quality checking, feature selection, and dataset splitting into training and testing data with a ratio of 70:30. From the initial 35 attributes, 16 main attributes were selected as the most relevant features to the target variable. The classification model was developed using the Decision Tree algorithm and evaluated using a confusion matrix, accuracy, precision, recall, and F1-score. The results showed that the model achieved an accuracy of 75.08% and a weighted average F1-score of 75.16%. Attributes such as Curricular units 2nd sem (approved), Curricular units 2nd sem (grade), and Tuition fees up to date were identified as the most influential factors in the classification process. Although the dataset had a class imbalance condition, this study maintained the original data distribution without applying oversampling techniques such as SMOTE in order to preserve the actual conditions of the scholarship selection process. In addition to providing fairly good classification performance, the Decision Tree method was also able to produce a transparent, interpretable, and easy-to-understand model through the resulting decision tree structure.