Classification of Student Grade Based on Academic Records Using Support Vector Machine

DOI

10.2991/assehr.k.201010.029How to use a DOI?

Keywords

Classification, Student’s Grade, Support Vector Machine (SVM), Synthetic Minority Oversampling Technique (SMOTE)

Abstract

Academic records have meaning “used” or “impressions” related to science in the form of notes. In this research, we use academic records of Statistics Students of Islamic University of Indonesia years 2015, which is like the percentage of late attendance, types of courses taken, schedule of days and hours of courses, and the number of SKS (Semester Credit System). This academic record is important because it has a pattern that affects the grade of the course. Therefore to find out the pattern of student academic records, it is necessary to classify the grade of the courses based on academic records, the Support Vector Machine (SVM) classification method is used because this method is reliable for classification with high dimensions and multiclass. In the academic record’s data it is known that there is an imbalance of data, so to overcome it, The Synthetic Minority Oversampling Technique (SMOTE) method are use SVM so the performance of classification would be better. We can conclude that by using the SVM and SMOTE method known that the classification has accuracy 58% with the Cost 10 and gamma 100 so that students who go in to “excellent” class are 363, “very good” class are 102, “good” class are 4, “fair” class are 2, “poor” class is only one, and “failed” class are 66.

Copyright

© 2020, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Amalia Dwi Nurfadzilah
AU  - Ayundyah Kesumawati
PY  - 2020
DA  - 2020/10/11
TI  - Classification of Student Grade Based on Academic Records Using Support Vector Machine
BT  - Proceedings of the 2nd International Seminar on Science and Technology (ISSTEC 2019)
PB  - Atlantis Press
SP  - 200
EP  - 206
SN  - 2352-5398
UR  - https://doi.org/10.2991/assehr.k.201010.029
DO  - 10.2991/assehr.k.201010.029
ID  - Nurfadzilah2020
ER  -