Building a Classification Model to Predict School Quality in Indonesia

DOI

10.2991/assehr.k.210423.074How to use a DOI?

Keywords

school quality, decision tree, random forest, boosted tree, classification

Abstract

The aim of the present study was to explore the best classification model based on the predictive performance of decision tree (DT), random forest (RF), and boosted tree (BT) models used for school quality status detection in Indonesia. In order to get a better insight into the predictive abilities of these three models, 213,536 records from Basic Education Data (Dapodik), an education management information system managed by Ministry of Education and Culture, were utilized. A total of 23 predictors were extracted from Dapodik. School quality status (two classes: ‘met national education standards’ and ‘improvement required’) was an output (dependent) variable. Accuracy of the DT, RF, and BT models evaluated on the training dataset and the testing dataset was 71.51%, 99.91%, and 72.66% as well as 71.53%, 73.56%, and 72.82%, respectively. The findings indicate that the BT model had better performance than the DT dan RF models.

Copyright

© 2021, 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  - Satriyo Wibowo
PY  - 2021
DA  - 2021/04/26
TI  - Building a Classification Model to Predict School Quality in Indonesia
BT  - Proceedings of the International Conference on Educational Assessment and Policy (ICEAP 2020)
PB  - Atlantis Press
SP  - 111
EP  - 114
SN  - 2352-5398
UR  - https://doi.org/10.2991/assehr.k.210423.074
DO  - 10.2991/assehr.k.210423.074
ID  - Wibowo2021
ER  -