Eye State Classification Using Ensemble Machine Learning Models and SMOTE on EEG Data

Electroencephalography (EEG) data presents complex and high-dimensional signals, offering great potential for applications in various fields such as neurofeedback, clinical diagnostics, cognitive neuroscience, human-computer interaction (HCI), and beyond. Analyzing EEG signals requires expertise not only from neuroscience, but also from signal processing, machine learning (ML), and statistics to extract meaningful information from brain activity recordings. Specifically, the combination of EEG and ML can provide an advantage in addressing challenging classification tasks in these fields. The present study focuses on the classification of eye state (open or closed) using Ensemble ML models such as Random Forest (RF), Gradient Boosting (GB), AdaBoost, XGBoost, and LightGBM on EEG data. We apply the Synthetic Minority Over-sampling Technique (SMOTE) to address the class imbalance and conduct a comparative analysis of the models’ performance with and without SMOTE using 10-fold cross-validation across several metrics namely, Accuracy, Precision, Recall, F1-score, and the Area Under the Curve (AUC). The experimental results highlight the importance of addressing the class imbalance in EEG data to improve model performance.