An ID3 Decision Tree Algorithm-Based Model for Predicting Student Performance Using Comprehensive Student Selection Data at Telkom University

Telkom University, in its routine admission process, generates a rich dataset consisting of various attributes of prospective students. These attributes extend beyond academic parameters like the grade point average (GPA) from the final high school year, encompassing non-academic data such as parent...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Ingénierie des systèmes d'Information 2023-10, Vol.28 (5), p.1205-1212
Hauptverfasser: Widaningsih, Sri, Muhamad, Wardani, Hendriyanto, Robbi, Nugroho, Heru
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Telkom University, in its routine admission process, generates a rich dataset consisting of various attributes of prospective students. These attributes extend beyond academic parameters like the grade point average (GPA) from the final high school year, encompassing non-academic data such as parental occupation, income, student's gender, origin province, high school major, and school category. Previous research has predominantly focused on academic and sociodemographic data, such as GPA and family income, respectively, for predicting study performance. However, factors like school major, study program, and school category have often been overlooked. In this study, the objective is to utilize the comprehensive Student Selection Data (SMB) to devise a model for predicting the performance of students in their first semester at Telkom University. The aim is to address the issue of a low rate of on-time graduation by leveraging the untapped potential of SMB data. An Iterative Dichotomiser 3 (ID3) decision tree algorithm forms the backbone of the proposed model, enabling the classification of student performance based on a range of diverse attributes. Information gain-based feature selection revealed the five attributes with the greatest influence on student performance in the first semester: gender, grade point average from the final year of high school, study program, high school major, and school category. These findings underscore the potential of a more inclusive approach to student data analysis in predicting academic success in higher education.
ISSN:1633-1311
2116-7125
DOI:10.18280/isi.280508