EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Machine Learning and Multi-Criteria Decision-Making Framework for Student Grade Prediction

Duško Tešić

Smart Multi-Criteria Analytics and Reasoning Technologies · 2025

DOI: 10.65069/smart1120253 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Smart Multi-Criteria Analytics and Reasoning Technologies, 2025. Verfasst von Duško Tešić. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper presents an integrated methodological framework that combines machine learning (ML) algorithms with multi-criteria decision-making (MCDM) methods to predict student grades based on multiple input criteria. Unlike traditional approaches that focus on assigning grades based on static thresholds, the proposed system allows for numerical prediction of overall achievement before formal evaluation, thereby providing prediction of educational outcomes and supporting necessary decision-making. The research used regression models, including Random Forest and Linear Regression, to model the relationships between input attributes and target variables. The obtained criterion weights from the models were used in the Simple Additive Weighting (SAW) method, which allowed for information aggregation and generation of proportional results. These results were then transformed into classification classes while preserving ranking and proportionality. Special emphasis is placed on technical reproducibility, interpretability and stability of the classification throughout all processing stages. Pearson correlation between predicted and scaled values, as well as between their ranks, confirms that the transformation preserved the data structure with high accuracy. This shows that MCDM methods can serve not only for evaluation and ranking, but also as a valid tool for numerical prediction in educational systems. The proposed framework enables the development of scalable, transparent and automated systems for predicting student grades, with the potential for wider application in educational analytics, instructional planning and identification of educational needs.

Zitieren

APA

Tešić, D. (2025). A Machine Learning and Multi-Criteria Decision-Making Framework for Student Grade Prediction. Smart Multi-Criteria Analytics and Reasoning Technologies. https://doi.org/10.65069/smart1120253

Vancouver

Tešić D. A Machine Learning and Multi-Criteria Decision-Making Framework for Student Grade Prediction. Smart Multi-Criteria Analytics and Reasoning Technologies. 2025. doi:10.65069/smart1120253

BibTeX
@article{Tei2025,
  author = {Tešić, Duško},
  title = {A Machine Learning and Multi-Criteria Decision-Making Framework for Student Grade Prediction},
  journal = {Smart Multi-Criteria Analytics and Reasoning Technologies},
  year = {2025},
  doi = {10.65069/smart1120253}
}
Kostenlos starten →