EUVIMEDEUVIMED entdecken
Crossref

Hybrid Deep Learning for EEG-Based Intelligence Prediction: Combining PSD, MSE, and CNN-GRU

Mohammed Alrowaili, Mohammed J. Alhaddad, Mutasem Jarrah, Saim Rasheed

International Journal of Advanced Computer Science and Applications · 2026 · 17 (10)

DOI: 10.14569/ijacsa.2026.0171090 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Advanced Computer Science and Applications, 2026. Verfasst von Mohammed Alrowaili und 3 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Electroencephalography (EEG) signals provide a non-invasive means of examining patterns of brain activity associated with cognitive processing. This study investigates whether spectral and complexity-based EEG characteristics can support the classification of intelligence quotient (IQ) levels using a hybrid deep-learning framework. EEG data were obtained from a publicly available dataset comprising 104 children aged 7–13 years who performed reading or mathematics tasks. Following data inspection, valid recordings from 90 participants were retained, yielding 212 task-level EEG samples for subsequent analysis. Because multiple task recordings could originate from the same participant, participant identity rather than individual task samples was considered when separating data for model evaluation. The EEG recordings were preprocessed and characterized using two complementary feature representations: Power Spectral Density (PSD), which describes frequency-domain activity, and Multiscale Entropy (MSE), which captures signal complexity across multiple scales. These representations were processed through a hybrid CNN–GRU architecture and used to classify participants into Low, Average, and High IQ categories. The proposed model achieved an overall classification accuracy of 95% in the reported evaluation and outperformed the conventional machine-learning and shallow neural-network models considered in the study. Class-specific precision, recall, and F1-score were also examined because of the unequal distribution of samples across the three IQ categories. The findings indicate that combining spectral and complexity-based EEG information may provide useful discriminative information for IQ-category classification within the investigated dataset. However, the relatively small number of independent participants, the pronounced class imbalance, the task-specific nature of the recordings, and the absence of external validation limit the extent to which these findings can be generalized. Further evaluation using larger and more diverse independent datasets is therefore needed to determine the robustness and broader applicability of the proposed framework.

Zitieren

APA

Alrowaili, M., Alhaddad, M. J., Jarrah, M., & Rasheed, S. (2026). Hybrid Deep Learning for EEG-Based Intelligence Prediction: Combining PSD, MSE, and CNN-GRU. International Journal of Advanced Computer Science and Applications, 17(10). https://doi.org/10.14569/ijacsa.2026.0171090

Vancouver

Alrowaili M, Alhaddad MJ, Jarrah M, Rasheed S. Hybrid Deep Learning for EEG-Based Intelligence Prediction: Combining PSD, MSE, and CNN-GRU. International Journal of Advanced Computer Science and Applications. 2026;17(10). doi:10.14569/ijacsa.2026.0171090

BibTeX
@article{Alrowaili2026,
  author = {Alrowaili, Mohammed and Alhaddad, Mohammed J. and Jarrah, Mutasem and Rasheed, Saim},
  title = {Hybrid Deep Learning for EEG-Based Intelligence Prediction: Combining PSD, MSE, and CNN-GRU},
  journal = {International Journal of Advanced Computer Science and Applications},
  year = {2026},
  volume = {17},
  number = {10},
  doi = {10.14569/ijacsa.2026.0171090}
}
Kostenlos starten →