<![CDATA[Electroencephalography (EEG) signals provide objective information on students’ cognitive responses; however, an appropriate classification method is required to effectively process their characteristics. This study aims to compare the performance of six machine learning algorithms in classifying students’ cognitive responses based on EEG data collected in the context of the Free Nutritious Meals (MBG) Program. The evaluated algorithms include K-Nearest Neighbors, Random Forest, Logistic Regression, Decision Tree, Naive Bayes, and Support Vector Machine, using delta, theta, alpha, beta, and gamma features. The modeling process involved baseline model development and hyperparameter tuning, followed by performance evaluation using accuracy, precision, recall, and F1-score. The results show that K-Nearest Neighbors achieved the highest baseline accuracy of 88.00%. After hyperparameter tuning, Decision Tree achieved the highest performance among the tuned models, with an accuracy and macro F1-score of 88.00%. However, its accuracy was identical to that of baseline KNN, and no significant difference was observed between their predictions. These findings indicate that hyperparameter tuning affected the evaluated algorithms differently, with Decision Tree showing the largest performance improvement under the experimental configuration used in this study.]]>
Copyrights © 2026
