Machine Learning-Based Teacher Performance Classification Using Administrative and Credit Point Assessment (PAK) Data: A Comparative Study of Decision Tree and Naive Bayes
Journal of Applied Informatics and Computing · 23.04.2026 · 10 (2) · 1853-1863
DOI: 10.30871/jaic.v10i2.12377 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Applied Informatics and Computing, online erschienen am 23.04.2026. Verfasst von MY Teguh Sulistyono und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
Teacher performance evaluation plays a strategic role in improving educational quality and supporting data-driven decision making. However, conventional evaluation approaches are often subjective and lack systematic data analysis. This study proposes a machine learning-based classification model to evaluate teacher performance using administrative data and Credit Point Assessment (PAK) records. The dataset consists of 30 teacher records with attributes including rank, functional position, and PAK score. Performance categories were derived based on predefined PAK score intervals, forming a structured rule-based classification framework. Data preprocessing included label encoding and min–max normalization, followed by an 80:20 stratified train–test split. Two supervised learning algorithms, Decision Tree and Naive Bayes, were implemented and evaluated using accuracy, precision, recall, F1-score, confusion matrix, and 5-fold cross-validation to ensure robustness. Experimental results show that both models achieved an accuracy of 0.83, with Decision Tree demonstrating more stable performance and higher interpretability. The relatively high performance is influenced by the deterministic structure of the PAK-based categorization and the limited dataset size. These findings indicate that Decision Tree is effective for automating structured administrative evaluation rules, while further validation using larger and multi-institutional datasets is necessary to improve generalizability. This study contributes to the development of transparent and data-driven teacher evaluation systems in educational institutions.
Zitieren
APA
Sulistyono, M. T., & Chaerunnisa, N. I. (2026). Machine Learning-Based Teacher Performance Classification Using Administrative and Credit Point Assessment (PAK) Data: A Comparative Study of Decision Tree and Naive Bayes. Journal of Applied Informatics and Computing, 10(2), 1853-1863. https://doi.org/10.30871/jaic.v10i2.12377
Vancouver
Sulistyono MT, Chaerunnisa NI. Machine Learning-Based Teacher Performance Classification Using Administrative and Credit Point Assessment (PAK) Data: A Comparative Study of Decision Tree and Naive Bayes. Journal of Applied Informatics and Computing. 2026;10(2):1853-1863. doi:10.30871/jaic.v10i2.12377
BibTeX
@article{Sulistyono2026,
author = {Sulistyono, MY Teguh and Chaerunnisa, Nur Ifani},
title = {Machine Learning-Based Teacher Performance Classification Using Administrative and Credit Point Assessment (PAK) Data: A Comparative Study of Decision Tree and Naive Bayes},
journal = {Journal of Applied Informatics and Computing},
year = {2026},
volume = {10},
number = {2},
pages = {1853--1863},
doi = {10.30871/jaic.v10i2.12377}
}
EUVIMED