AccScience Publishing / IJOCTA / Online First / DOI: 10.36922/IJOCTA026110037
Submit to IJOCTA
Cite this article
5
Download
41
Views
Journal Browser
Volume | Year
Issue
Search
Forthcoming Issue
Current Issue
View All
News and Announcements
View All
REVIEW ARTICLE

A comparative study of deep neural network architectures with chaotic optimization for electroencephalography-based emotion recognition

Zahraa Tarek1 ,  Khaled Alnowaiser2 ,  Esraa Hassan3*
Show Less
1 Department of Computer Engineering and Information, College of Engineering, Wadi Ad Dwaser, Prince Sattam Bin Abdulaziz University, Al-Kharj, Saudi Arabia
2 Department of Computer Science, College of Computer Engineering and Sciences, Prince Sattam Bin Abdulaziz University, Al-Kharj, Riyadh, Saudi Arabia
3 Department of Machine Learning and Information Retrieval, Faculty of Artificial Intelligence, Kafrelsheikh University, Kafr El Sheikh, Kafr El Sheikh, Egypt
IJOCTA, 026110037 https://doi.org/10.36922/IJOCTA026110037
Received: 10 March 2026 | Revised: 20 April 2026 | Accepted: 23 April 2026 | Published online: 21 May 2026
© 2026 by the Author(s). This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution -Noncommercial 4.0 International License (CC-by the license) ( https://creativecommons.org/licenses/by-nc/4.0/ )
Download PDF
XML
Cite
Abstract

Electroencephalography (EEG)-based emotion recognition has become an important research direction in affective computing, driven by its applications in mental health assessment, human--computer interaction, and brain--machine interfaces. Despite recent progress, challenges persist due to the high dimensionality of EEG features and the absence of consensus on optimal deep learning architectures for reliable emotion classification. This study introduces a unified comparative framework for EEG emotion recognition that combines statistical feature extraction with chaos-enhanced deep learning optimization. Five representative neural architectures, Transformer-based, CNN--LSTM hybrid, residual multilayer perceptron (MLP), capsule network, and hyperbolic-inspired network, were systematically evaluated under identical experimental conditions. EEG signals were normalized and transformed using advanced statistical feature extraction techniques, followed by robust data augmentation to improve model generalization. A chaos-driven learning rate optimization strategy based on the logistic map was incorporated into the training process to promote stable convergence and reduce susceptibility to local minima. Experimental results demonstrated that chaotic optimization substantially improved training stability and overall classification performance across architectures. The hyperbolic-inspired network achieved the highest validation accuracy of 93.21\% with a validation loss of 0.197 and a training time of 19.09 s, followed closely by the residual MLP, which attained 91.80\% accuracy with a loss of 0.226. The capsule network also showed competitive performance (90.87\% accuracy), whereas the CNN--LSTM hybrid exhibited comparatively lower accuracy (70.73%). These findings underscore the effectiveness of chaos-driven optimization and provide practical insights into architecture selection for efficient and robust EEG-based emotion recognition. The proposed framework offers a reproducible benchmark to support informed model design in affective computing applications.

Keywords
Cloud-based systems
Failure prediction
Random forest
CatBoost
Light gradient boosting machine
Principal component analysis
Likelihood of failure
Funding
The authors extend their appreciation to Prince Sattam bin Abdulaziz University for funding this research work through the project number (PSAU/2025/01/36818)
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References
  1. Li C, Zhang Z, Song R, Cheng J, Liu Y, Chen EEG-based emotion recognition via neural architecture search. IEEE Trans Affect Comput. 2021;14(2):957-968. https://www.doi.org/10.1109/TAFFC.2021.3130387
  2. Hasan R, Islam A comparative analy-sis of emotion recognition from EEG signals using temporal features and hyperparameter-tuned machine learning techniques. MethodsX. 2025:103468. https://www.doi.org/10.1016/j.mex.2025.103468
  3. Keelawat P, Thammasan N, Numao M, Kijsirikul A comparative study of window size and chan-nel arrangement on EEG-emotion recognition us-ing deep CNN. Sensors (Basel). 2021;21(5):1678. https://www.doi.org/10.3390/s21051678.
  4. Joloudari JH, Maftoun M, Nakisa B, Alizadehsani R, Yadollahzadeh-Tabari Complex emotion recognition system using basic emotions via fa-cial expression, EEG, and ECG signals: a review. Preprint. arXiv. 2024. https://www.doi.org/10.48550/arXiv.2409.07493
  5. Kumari N, Anwar S, Bhattacharjee A com-parative analysis of machine and deep learning techniques for EEG evoked emotion classification. Wirel Pers Commun. 2023;128(4):2869-2890. https://www.doi.org/10.1007/s11277-022-10076-7
  6. Haider Cooperative chameleon search opti-mization enabled deep ensemble classifier for emo-tion recognition using EEG signal. Comput Meth-ods Biomech Biomed Eng. 2025:1-23. https://www.doi.org/10.1080/10255842.2025. 2568979
  7. Yildirim E, Kaya Y, Kilic¸ F. A channel selection method for emotion recognition from EEG based on swarm-intelligence algorithms. IEEE Access. 2021;9:109889-109902. https://www.doi.org/10.1109/ACCESS.2021. 3100638
  8. Guo X, Liang R, Xu S, Dong L, Liu Y. An in-vestigation of echo state network for EEG-based emotion recognition with deep neural networks. Biomed Signal Process Control. 2026;111:108342. https://www.doi.org/10.1016/j.bspc.2025.108342
  9. Jehosheba Margaret M, Masoodhu Banu NM. Per-formance analysis of EEG-based emotion recogni-tion using deep learning models. Brain Comput Interfaces. 2023;10(2-4):79-98. https://www.doi.org/10.1080/2326263X.2023.
  10. Muniyandi  AP,  Padmanandam  K,  Subbaraj  K, et    An  intelligent  emotion  prediction  system using improved sand cat optimization technique based on EEG signals. Sci Rep. 2025;15(1):8782. https://www.doi.org/10.1038/s41598-025-89904-2
  11. Huang H, Deng Y, Hao B, Liu W, Tu X, Zeng Emotion recognition method using U-Net neu-ral network with multichannel EEG features and differential entropy characteristics. IEEE Access. 2024. https://www.doi.org/10.1109/ACCESS.2024. 3497160
  12. Mir WA, Anjum M, Shahab S. Deep-EEG: an optimized and robust framework and method for EEG-based diagnosis of epileptic Diag-nostics (Basel). 2023;13(4):773. https://www.doi.org/10.3390/diagnostics 13040773
  13. Zhang S, Ling C, Wu J, et al. EEG-ERnet: emo-tion recognition based on rhythmic EEG convo-lutional neural network model. J Integr Neurosci. 2025;24(8):41547. https://www.doi.org/10.31083/JIN41547
  14. Maria MA, Akhand MAH, Hossain AA, Kamal MAS, Yamada A comparative study on promi-nent connectivity features for emotion recognition from EEG. IEEE Access. 2023;11:37809-37831. https://www.doi.org/10.1038/s41598-023-40786-2
  15. Zheng W, Peng Y, Zhang A, Yuan EEG-based emotion identification from nerve conduc-tion mechanisms: a gustatory-emotion coupling model combined with multiblock attention mod-ule. Expert Syst Appl. 2026;298:129855. https://www.doi.org/10.1016/j.eswa.2025.129855
  16. Wu Y, Lu, L, Xu A, et Neural networks for epilepsy detection and prediction with EEG signals: a systematic review. Artif Intell Rev. 2025;59(1):30. https://www.doi.org/10.1007/s10462-025-11441-1
  17. C¸ okc¸etin B, Uc¸ar MK. A PRISMA-based system-atic  review  on  advances  in  identity  recognition and authentication using human biometric signals (2018–2023). Biomed Eng Online. https://www.doi.org/10.1186/s12938-025-01508-z
  18. Wang R, Xu H, Ma Y, Che Research on the classification  of  EEG  signals  for  dementia  and its interpretability using the GWOCS algorithm. Cogn Neurodyn. 2025;20(1):1. https://www.doi.org/10.1007/s11571-025-10348-5
  19. Xu X, Chen C, Sun Z, et al. Research on control strategy of lower limb exoskeleton robots: a review. Sensors (Basel). 2026;26(2):255. https://www.doi.org/10.3390/s26020355
  20. Yu F, Liu JM. A deep learning-based framework for sentiment and emotion classification of so-cial media messages during pandemic periods. J Circuits Syst Comput. 2025;35(3):2550413. https://www.doi.org/10.1142/S0218126625504134
  21. Eisa MM, Alnaggar MH. Hybrid rough-genetic classification model for IoT heart disease monitor-ing system. In: Magdi DA, Helmy YK, Mamdouh M, Joshi A, eds. Digital Transformation Technol-ogy. Lecture Notes in Networks and Systems, vol Springer. 2022:437-451. https://www.doi.org/10.1007/978-981-16-2275-5 27
  22. Abdella MH, Hawash AA, Zahran OE, et Smart tour guide: a novel artificial intelligence system for replacing human guides in cultural heritage sites. Egypt J Artif Intell. 2024;3(1). https://www.doi.org/10.21608/ejai.2024.226751.1017
  23. Hassan E, Bhatnagar R, Shams MY. Advancing scientific research in computer science by Chat-GPT and LLaMA–A review. In: Sai PHVST, Pot-nuru S, Avcar M, Kar VR, eds. Intelligent Man-ufacturing and Energy Sustainability. Springer; 2023:23-37. https://www.doi.org/10.1007/978-981-99-6774-2 3
  24. Karthiga M, Suganya E, Sountharrajan S, et al. EEG-based smart emotion recognition using meta heuristic optimization and hybrid deep learning tec Sci Rep. 2024;14(1):30251. https://www.doi.org/10.1038/s41598-024-80448-5
  25. Ibrahim MA, Ali ME, Ahmed MN, et A low-cost approach for drowning detection and alert-ness. Egypt J Artif Intell. 2024;3(1). https://www.doi.org/10.21608/ejai.2024.223796.1012
  • Davarzani S, Masihi S, Panahi M, et al. A com-parative study on machine learning methods for EEG-based human emotion Electron-ics (Basel). 2025;14(14):2744. https://www.doi.org/10.3390/electronics 14142744
  • Hassan E, El-Rashidy N, Elbedwehy S, et al. Ex-ploring the frontiers of image super-resolution: a review of modern techniques and emerging appli-cations. Neural Comput Appl. 2025;37(22):17913- https://www.doi.org/10.1007/s00521-025-11331-1
  1. Hamzah HA, Abdalla KK. EEG-based emotion recognition systems: comprehensive study. He-liyon. 2024;10(10). https://www.doi.org/10.1016/j.heliyon.2024. e31485
  2. Kargarnovin S, Hernandez C, Farahani FV, Kar-wowski Evidence of chaos in electroencephalo-gram signatures of human performance: a system-atic review. Brain Sci. 2023;13(5):813. https://www.doi.org/10.3390/brainsci13050813
  3. Hassan E, Bhatnagar R, Abd El-Hafeez T, Shams MY. Detection of suicide and depression for early intervention and initiative-taking mental health-care. In: Proc IEEE ISPCC. 2025:99-104. https://www.doi.org/10.1109/ISPCC66872.2025. 11039547
  4. Alnaggar MH, El-Dosuky M, Rashad M. A novel multimodal biometric template security based on nano-scale reaction-diffusion J Comput Theor Nanosci. 2018;15(6-7):1979-1982. https://www.doi.org/10.1166/jctn.2018.7522
  5. Chhabra H, Vempati R, Chauhan U, et al. Au-tomated human emotion recognition from EEG signals using chaotic local binary pattern and ensemble learning. Int J Mach Learn Cybern. 2026;17(1):12. https://www.doi.org/10.1007/s13042-025-02822-7
  6. Hasoun RK. Hybrid optimized feature selection and deep learning method for emotion recogni-tion that uses EEG data. Iraqi J Comput Inform. 2025;51(1):1-18. https://www.doi.org/10.25195/ijci.v51i1.545
  7. Shams MY, Hassan E, Gamil S, et al. Skin dis-ease classification: a comparison of ResNet50, Mo-bileNet, and Efficient-B0. J Curr Multidiscip Res. 2025;1(1):1-7. https://www.doi.org/10.21608/jcmr.2025.327880. 1002
  8. Flower TML, Singh SCE, Jaya T, Devadhas GG. EEG-based emotion recognition: a systematic re-view of traditional and deep learning methods. Science. 2025;15:100180.

https://www.doi.org/10.70389/PJS.100180

  1. Tan S, Tang Z, He Q, et al. Automatic detection and prediction of epileptic EEG signals based on nonlinear dynamics and deep learning: a review. Front Neurosci. 2025;19:1630664. https://www.doi.org/10.3389/fnins.2025.1630664
  2. Kumar BA, Suresh HN, Ranjitha S. EEG-based assessment of suicidality risk: an integrated frame-work with self-adaptive chaotic cuckoo search and Eng Technol Appl Sci Res. 2026;16(1):32391-32397. https://www.doi.org/10.48084/etasr.14247
  3. Saber A, Hassan E, Elbedwehy S, Awad WA, Emara TZ. Leveraging ensemble convolutional neural networks and metaheuristic strategies for advanced kidney disease screening and classifica-tion. Sci Rep. 2025;15(1):12487. https://www.doi.org/10.1038/s41598-025-93950-1
  4. Al-Qammaz An enhanced computational model based on social spider optimization algo-rithm for EEG-based emotion recognition. Disser-tation. Universiti Utara Malaysia; 2019. https://www.doi.org/10.14419/ijet.v7i2.15.11373
  5. Hassan E, Saber A, Alqahtani O, El-Rashidy N, Elbedwehy S. An innovative approach to advanced voice classification of sacred Quranic recitations through multimodal fusion. Egypt In-form J. 2025;30:100640. https://www.doi.org/10.1016/j.eij.2025.100640
Next article in this issue
Share
Back to top
An International Journal of Optimization and Control: Theories & Applications, Electronic ISSN: 2146-5703 Print ISSN: 2146-0957, Published by AccScience Publishing
We use cookies on our website to ensure you get the best experience.
Read more about our cookies here
Accept