Comparison of Machine Learning and Deep Learning Models for Mental Condition Classification Based on Electroencephalogram Data
Likelihood: Journal of Statistics and Its Application · 31.12.2025 · 1 (2) · 87-95
DOI: 10.54065/likelihood.1245 ↗Kurz gefasst
Fachartikel in der Zeitschrift Likelihood: Journal of Statistics and Its Application, online erschienen am 31.12.2025. Verfasst von Saffanah Karimah. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Mental states refer to the state of mind that can be viewed from various perspectives, such as consciousness-based, intention-based, and functionalism-based. Among the various types of human mental states, those that are inherent in everyday life are concentration, relaxation, and a neutral state between the two. Concentration is a state that needs to be achieved in activities related to human cognitive abilities, while relaxation is necessary when not performing strenuous activities so that the body can feel comfortable. Therefore, the ability to achieve these two states when needed is very important. However, in reality, people with mental disorders or patients with neurological diseases often find it difficult to achieve the ideal state of concentration or relaxation. Therefore, immediate treatment is needed to overcome the existing effects. Treatment of patients often requires an assessment of their current mental condition. One method that can be used to detect mental conditions is to use electroencephalogram (EEG) brain wave signals. Understanding EEG patterns can provide a more objective picture of the body's condition because every neurological and mental condition in humans affects the brain wave patterns that appear. Brain wave signals have a complex structure and are rich in information. The application of machine learning and deep learning algorithms is considered capable of learning and recognizing patterns in EEGs so that characteristics that distinguish the patterns of each mental condition can be identified. In this study, by comparing the machine learning method, namely Multiclass Support Vector Machine (SVM) using the One-Against-One approach, and the deep learning method, namely Artificial Neural Network (ANN), it was found that the ANN method had a higher AUC value of 0.878 compared to the Multiclass SVM method with an AUC of 0.767. Therefore, the ANN method with Stochastic Gradient Descent optimization is the best method for classifying mental states of concentration, neutrality, and relaxation based on EEG data.
Zitieren
APA
Karimah, S. (2025). Comparison of Machine Learning and Deep Learning Models for Mental Condition Classification Based on Electroencephalogram Data. Likelihood: Journal of Statistics and Its Application, 1(2), 87-95. https://doi.org/10.54065/likelihood.1245
Vancouver
Karimah S. Comparison of Machine Learning and Deep Learning Models for Mental Condition Classification Based on Electroencephalogram Data. Likelihood: Journal of Statistics and Its Application. 2025;1(2):87-95. doi:10.54065/likelihood.1245
BibTeX
@article{Karimah2025,
author = {Karimah, Saffanah},
title = {Comparison of Machine Learning and Deep Learning Models for Mental Condition Classification Based on Electroencephalogram Data},
journal = {Likelihood: Journal of Statistics and Its Application},
year = {2025},
volume = {1},
number = {2},
pages = {87--95},
doi = {10.54065/likelihood.1245}
}
EUVIMED