Machine learning (ML) approaches have been widely used in bio-signal analysis and disease classification. ML techniques have been used in emotion recognition [29] and the prediction of diseases including dementia [30], stroke [31], and AD [32]. Furthermore, ML techniques have been used on EEG signals to understand their complex electrophysiological activities and characterize the dynamic features of a complex brain network. Several studies have utilized traditional machine learning models such as the K-nearest neighbor (KNN), decision tree (DT), random forest (RF), Naïve Bayes, and regression models to investigate neurological disorders [33,34]. In a recent study, support vector machines (SVM), KNN, and Naïve Bayes were used to predict the AD [32] and SVM, correctly classifying 83% of the subjects using network features. To classify healthy aging EEG signals using network features, SVM achieved an accuracy above 80% [6]. A low-density device with seven electrodes was designed for an automated EEG-based AD detection system, and the SVM obtained 91.1% accuracy [35]. A portable EEG device was used to quantify the mental workload during the driving and the binary machine learning models achieved
