Classification and forecasting in students' progress using Multiple-Criteria Decision Making, K-Nearest Neighbors, and Multilayer Perceptron methods

Slađana Spasić1   and Violeta Timašević2  

  1. University of Belgrade - Institute for Multidisciplinary Research
    Kneza Višeslava 1, 11030 Belgrade, Serbia
    sladjana@imsi.bg.ac.rs
  2. Singidunum University, Faculty of Informatics and Computing
    Danijelova 32, 11000 Belgrade, Serbia
    vitomasevic@singidunum.ac.rs

Abstract

The research paper addresses students' performance in higher education. It proposes using the MCDM method - Promethee II to assess students' knowledge and the K-Nearest Neighbors (KNN) and Multilayer Perceptron (MLP) methods for grade classification. The main goals are tracking and diagnosing students' knowledge levels, predicting their outcomes, and providing tailored recommendations. It helps to identify students at risk of not passing the course and evaluates teaching methods. This encourages student engagement and progress during the course. The research demonstrates the suitability of Promethee II, MLP, and KNN methods for effectively monitoring, classifying, and predicting students' progress during the semester, enhancing the objectivity of the assessment process.

Key words

Promethee II, MLP, KNN, student's grades mark classification, student's achievement forecasting, Matthews Correlation Coefficient, Class Balance Accuracy

Digital Object Identifier (DOI)

https://doi.org/10.2298/CSIS241022050S

Publication information

Volume 22, Issue 3 (June 2025)
Year of Publication: 2025
ISSN: 2406-1018 (Online)
Publisher: ComSIS Consortium

Full text

DownloadAvailable in PDF
Portable Document Format

How to cite

Spasić, S., Timašević, V.: Classification and forecasting in students' progress using Multiple-Criteria Decision Making, K-Nearest Neighbors, and Multilayer Perceptron methods. Computer Science and Information Systems, Vol. 22, No. 3, 881–905. (2025), https://doi.org/10.2298/CSIS241022050S