O
Orclever
Back to Journal
Research Article Open AccessOrclever Native

EEG-Based Assessment of Stress Levels Using Time–Frequency Features and Machine Learning

Sevde Samsa1,
Çiğdem Gülüzar Altıntop2✉
1Erciyes University, Graduate School of Natural and Applied Sciences, Biomedical Engineering Graduate Program, Kayseri, Türkiye
2Erciyes University, Faculty of Engineering, Department of Biomedical Engineering, Kayseri, Türkiye
Received:Nov 5, 2025→Revised:Feb 20, 2026→Accepted:Mar 23, 2026→Published:March 25, 2026
DOI: 10.56038/ejrnd2026297531
Vol. 6, No. 1 · ejrnd2026297531
4 Views2 Downloads0 Cited by
AI SummaryAI-generated · verify against the full text

EEG signals can be used to accurately classify stress levels using machine learning techniques and a limited number of brain channels.

Researchers used the Empirical Wavelet Transform to break down EEG signals into smaller parts and then applied machine learning algorithms to classify stress levels. They found that using a Random Forest classifier with a limited number of EEG channels was effective in identifying stress levels, with an accuracy of 86.25% for binary classification. The most useful features for classification were related to signal power, relative energy, and amplitude, particularly from frontal and temporal brain channels.

Abstract

Stress from everyday living and exam times can have a major impact on cognitive function. So, identifying the degree of stress can be crucial in choosing preventive measures. In order to categorize stress levels (binary and multi-class) using machine learning techniques, the Empirical Wavelet Transform was used in this study to decompose EEG signals into subbands. Accuracy, sensitivity, specificity, precision, and F-score measurements were used to evaluate the performance of machine learning algorithms. The findings demonstrated that binary classification yielded higher accuracy than other classification tasks, especially between low and high stress levels. After feature selection, the Random Forest classifier produced the best results, with an accuracy of 86.25%. The K-Nearest Neighbor algorithm produced the best accuracy of 66.67% in multi-class classification, demonstrating the greater challenge of differentiating stress levels. The most instructive features, as determined by statistical analysis, were associated with signal power, relative energy, and amplitude, which were mostly obtained from frontal and temporal EEG channels. This study demonstrates that stress levels can be effectively classified using a limited number of EEG channels and simple features, providing a practical approach for EEG-based stress assessment.
Keywords
ElectroencephalogramStress levelsBiomedical signal processingMachine learningFeature extractionStress detection

Full TextOpen Access

References

  1. 1.[1] S. Das, S. Chatterjee, A. I. Karani, and A. K. Ghosh, "Stress detection while doing exam using eeg with machine learning techniques," in Proc. Int. Conf. Innov. Data Anal. (ICIDA), Springer, 2023, pp. 177–187.
  2. 2.[2] H. M. Afify, K. K. Mohammed, and A. E. Hassanien, "Stress detection based EEG under varying cognitive tasks using convolution neural network," Neural Comput. Appl., vol. 37, pp. 5381–5395, 2025.
  3. 3.[3] A. Gandhi and K. Udesang, "Stress detection through EEG signals: employing a hybrid approach integrating time domain, frequency domain features and machine learning techniques," J. Electr. Syst., vol. 20, no. 4, pp. 3965–3973, 2024.
  4. 4.[4] Y. Badr, U. Tariq, F. Al-Shargie, F. Babiloni, F. Al Mughairbi, and H. Al-Nashash, "A review on evaluating mental stress by deep learning using EEG signals," *Neural Comput. Appl.*, vol. 36, pp. 12629–12654, 2024.
  5. 5.[5] B. Roy, L. Malviya, R. Kumar, S. Mal, A. Kumar, T. Bhowmik, and et al., "Hybrid deep learning approach for stress detection using decomposed EEG signals," *Diagnostics*, vol. 13, no. 10, p. 1936, 2023.
  6. 6.[6] A. Hag, F. Al-Shargie, D. Handayani, and H. Asadi, "Mental stress classification based on selected electroencephalography channels using correlation coefficient of Hjorth parameters," *Brain Sci.*, vol. 13, no. 1, p. 1340, 2023.
  7. 7.[7] A. Siripongpan, T. Namkunee, P. Uthansakul, T. Jumphoo, and P. Duangmanee, "Stress among Medical Students Presented with an EEG at Suranaree  University of Technology, Thailand ," Health Psychology Research, vol. 10, no. 2, pp. 1, 2022. doi: 10.52965/001c.35462.
  8. 8.[8] M. Rahman Momo, Md. Tahsin, A. Hossain, R. Shikder, M. Hossain Khan, R. U. Islam, and MR. A. Rashid, "A Comprehensive Dataset of EEG Recordings Capturing Student Stress Responses During Exams," Mendeley Data, vol. V1, 2025. doi: 10.17632/fyj9by2t22.1.
  9. 9.[9] G. J. Gilles, "Empirical wavelet transform," *IEEE Trans. Signal Process.*, vol. 61, pp. 3999–4010, 2013.
  10. 10.[10] Ç. G. Altıntop, F. Latifoğlu, A. K. Akın, and A. Ülgey, "Quantitative electroencephalography analysis for improved assessment of consciousness levels in deep coma patients using a proposed stimulus stage," *Diagnostics*, vol. 13, no. 8, p. 1383, 2023.
  11. 11.[11] E. Uğurgöl, M. Altınkaynak, D. Yeşilbaş, T. Batbat, A. Güven, E. Demirci, and et al., "Investigating the neural correlates of stroop effect using the multilayer perceptron neural network," Exp. Biomed. Res., vol. 7, 2024.
  12. 12.[12] P. Chawla, S. B. Rana, H. Kaur, K. Singh, R. Yuvaraj, and M. Murugappan, "A decision support system for automated diagnosis of Parkinson’s disease from EEG using FAWT and entropy features," Biomedical Signal Processing and Control, vol. 79, pp. 104116, 2023. doi: 10.1016/j.bspc.2022.104116.
  13. 13.[13] Ç. G. Altıntop, F. Latifoğlu, and A. K. Akın, "Can patients in deep coma hear us? Examination of coma depth using physiological signals," Biomedical Signal Processing and Control, vol. 77, pp. 103756, 2022. doi: 10.1016/j.bspc.2022.103756.
  14. 14.[14] J. Gou, H. Ma, W. Ou, S. Zeng, Y. Rao, and H. Yang, "A generalized mean distance-based k-nearest neighbor classifier," Expert Systems with Applications, vol. 115, pp. 356-372, 2019. doi: 10.1016/j.eswa.2018.08.021.
  15. 15.[15] C. Cortes and V. Vapnik, "Support-vector networks," Machine Learning, vol. 20, no. 3, pp. 273-297, 1995. doi: 10.1007/bf00994018.
  16. 16.[16] B. Yegnanarayana, Artificial Neural Networks. New Delhi, India: PHI Learning Pvt. Ltd., 2009.
  17. 17.[17] Ç. G. Altıntop, "Beyond Conventional Blood Parameters: Novel Hematologic Indices for Interpretable Artificial Intelligence in Acute Myocardial Infarction," *J. Clin. Pr. Res.*, vol. 47, p. 0, 2025.
  18. 18.[18] Margineantu DD, Dietterich TG. Pruning adaptive boosting. ICML, vol. 97, 1997, p. 211–8.
  19. 19.[19] Y. Dong, L. Xu, J. Zheng, D. Wu, H. Li, Y. Shao, and Y. Shao, "A Hybrid EEG-Based Stress State Classification Model Using Multi-Domain Transfer Entropy and PCANet," *Brain Sci.*, vol. 14, no. 1, p. 595, 2024.
  20. 20.[20] M. Mynoddin, T. Dev, and R. Chakma, "Brain2Vec: A Deep Learning Framework for EEG-Based Stress Detection Using CNN-LSTM-Attention," *arXiv Prepr. arXiv250611179*, 2025.
  21. 21.[21] J. J. Gonzalez-Vazquez, L. Bernat, J. L. Ramon, V. Morell, and A. Ubeda, "A Deep Learning Approach to Estimate Multi-Level Mental Stress From EEG Using Serious Games," IEEE Journal of Biomedical and Health Informatics, vol. 28, no. 7, pp. 3965-3972, 2024. doi: 10.1109/jbhi.2024.3395548.
  22. 22.[22] A. Hag, D. Handayani, T. Pillai, T. Mantoro, M. H. Kit, and F. Al-Shargie, "EEG mental stress assessment using hybrid multi-domain feature sets of functional connectivity network and time-frequency features," *Sensors*, vol. 21, no. 17, p. 6300, 2021. doi: 10.3390/s21176300.

Contents

5/53,994 w
Download PDF
Cite This Article
Samsa, S., Altıntop, Ç. G. (2026). EEG-Based Assessment of Stress Levels Using Time–Frequency Features and Machine Learning. *The European Journal of Research and Development*, 6(1), ejrnd2026297531. https://doi.org/10.56038/ejrnd2026297531

Bibliographic Info

JournalThe European Journal of Research and Development
Volume6
Issue1
Pages1–11
Article IDejrnd2026297531
PublishedMarch 25, 2026
eISSN2822-2296

Indexing & License

Open AccessCC BY 4.0CrossRef DOIORCIDOAI-PMH

More from The European Journal of Research and Development

Challenges in Maize Root Phenotyping: Preprocessing Limits and Class Imbalance in Deep Learning

Hüdanur Engin, Ali Murat Tiryaki

2026 · Vol 6 · Issue 1

The Bleaching of Woven Fabrics Using the Foam Application Technique

Aylin Kuşen, Onur Balcı, Koray Pektaş

2026 · Vol 6 · Issue 1

Automated Monkeypox Disease Classification Using Texture and Focus-Based Image Features

Tuğba Şentürk, Çiğdem Gülüzar Altıntop, Fatma Latifoğlu

2026 · Vol 6 · Issue 1

A Compact Non-Intrusive Measurement System for Critical Dimensions and Calibration Chart Generation of Underground Fuel Tanks

İlker Değirmencioğlu, Savaş Barış, Yusuf Kaya

2025 · Vol 5 · Issue 1

An AI-Based Question–Answering System for Corporate Documents: VK ArtiFin

Zeynep Örpek, Büşra Tural, Zeynep Destan

2025 · Vol 5 · Issue 1

EARS-XTSK: Privacy-Preserving Global Explainability in Cross-Silo Federated Two-Tower Recommendation Systems via Server-Side TSK Fuzzy Rule Distribution

Deniz Altay Avcı

2025 · Vol 5 · Issue 1

Browse all articles in The European Journal of Research and Development →