Regional Brain Analysis and Machine Learning Techniques for Classifying Familiar and Unfamiliar Faces Using EEG

Among the processes critical for human social interactions are perception, memorization, and bonding, and the ability to recognize familiar and unfamiliar faces is one of the most essential aspects of the human brain. This is a valuable communication skill, as well as remembering people and interpersonal interactions to recognize the faces of the people. Given the importance of these aspects of cognitive functioning, the present research seeks to establish the neural basis for recognizing familiar and unfamiliar faces from EEG data through a regional brain perspective and simple neural networks. The EEG data used in this research were gathered from typically developed subjects, and the features were derived using discrete wavelet transform (DWT). These features were then employed for the classification of the network using three different algorithms, which include k-nearest neighbors (KNN), support vector machines (SVM), and probabilistic neural networks (PNN). KNN was seen to have the highest classification accuracy than the other classifiers; the accuracy was considered for different brain regions and all the channels. The temporal and occipital lobes were found to be involved in face recognition, and the patterns of activation differed between familiar and unfamiliar faces. This work contributes to the literature by describing how face recognition is implemented in the brain, which areas of the brain are most important, and by comparing machine learning techniques for classifying the EEG signal. These findings are helpful for the current literature. They can help to inform future research into the neural structure of face recognition and what this might mean for prosopagnosia and similar conditions.