Machine Learning-Based Detection of Student Stress Levels Through Socioeconomic Status
University of Essex · 18.02.2026
DOI: 10.5526/esj.453 ↗Kurz gefasst
Fachartikel in der Zeitschrift University of Essex, erschienen am 18.02.2026. Verfasst von Tapu Biswas und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Stress among university students is a significant issue affecting their emotional, physical, and cognitive well-being. Stress which can lead to anxiety, depression, and impaired academic performance. Early detection is important for mental health and academic success. This study used machine learning to predict stress levels in undergraduate students based on socioeconomic data from an online survey in Bangladesh, using the PSS-10 stress scale. Results showed varying stress levels among students. Ten models were tested—including Naive Bayes, Random Forest, Decision Tree, CART, Logistic Regression, K-Nearest Neighbors, SVM, AdaBoost, and Bagging with the highest accuracy of 97.92% achieved by Decision Tree, CART, AdaBoost, and Bagging.
Schlagwörter
Zitieren
APA
Biswas, T., Mim, R. K., & Ferdous, F. S. (2026). Machine Learning-Based Detection of Student Stress Levels Through Socioeconomic Status. University of Essex. https://doi.org/10.5526/esj.453
Vancouver
Biswas T, Mim RK, Ferdous FS. Machine Learning-Based Detection of Student Stress Levels Through Socioeconomic Status. University of Essex. 2026. doi:10.5526/esj.453
BibTeX
@article{Biswas2026,
author = {Biswas, Tapu and Mim, Razowana Khan and Ferdous, Farhan Sadik},
title = {Machine Learning-Based Detection of Student Stress Levels Through Socioeconomic Status},
journal = {University of Essex},
year = {2026},
doi = {10.5526/esj.453}
}
EUVIMED