| M. Trigka, E. Dritsas, Ph. Mylonas |
| Sleepiness Detection Using Machine Learning Models on EEG Data |
| 13th Conference on Artificial Intelligence (SETN 2024), September 11-13, 2024, Piraeus, Greece |
|
ABSTRACT
|
| Driver sleepiness is a major cause of road accidents, necessitating effective detection systems to improve safety. This study investigates the use of machine learning (ML) models to automate the detection of driver sleepiness through electroencephalography (EEG) data collected in simulated environments. Various ML models, such as Random Forests (RF), Decision Trees (DT), Logistic Model Trees (LMT) and two ensemble methods (bagging and stacking), were evaluated using 10-fold cross-validation. More specifically, the selected classifiers were trained and tested using EEG data acquired via the MindSet device, including band power, attention, and mediation features to effectively differentiate between "sleepy" and "non-sleepy" subjects. The bagging approach demonstrated superior performance among the classifiers, achieving 74.9% accuracy, 0.749 precision, 0.750 recall, and an Area Under the ROC Curve (AUC) of 0.835.
|
| 11 September, 2024 |
| M. Trigka, E. Dritsas, Ph. Mylonas, "Sleepiness Detection Using Machine Learning Models on EEG Data", 13th Conference on Artificial Intelligence (SETN 2024), September 11-13, 2024, Piraeus, Greece |
[  PDF] [
BibTex] [
Print] [
 Back] |
