EEG-Based Biometric Authentication by Using Auditory Evoked Potential: Evaluation of Classifier Performance
摘要
Biometric authentication using EEG signals, specifically auditory evoked potentials (AEP), offers robust security advantages over traditional methods. This paper evaluates the performance of various classifiers—Support Vector Machine (SVM), K-Nearest Neighbour (KNN), and Random Forest (RF) in EEG-based biometric systems. Employing Shannon entropy and Linear Discriminant Analysis (LDA) for feature extraction and selection, respectively, we found SVM achieved the highest classification accuracy (96.15%), precision, recall, and F1-score. These results demonstrate SVM’s effectiveness for EEG-based authentication systems.