“…For the aim of finding the relative EEG markers that explain mental stress and increase its detection rate, several studies employed different types of features from the time domain, frequency domain, and time-frequency domain [ 8 , 32 , 33 , 34 , 35 , 36 ], and several machine learning algorithms have been used to predict the mental stress state, such as SVM [ 37 ], K-Nearest Neighbors(KNN) [ 29 , 38 ], LR [ 1 ], Feed-Forward Neural Network (FF-NN) [ 30 ], Naive Bayes(NB) [ 9 , 38 ], and Random Forest(RF) [ 39 ]. In the literature, non-invasive EEG-based stress studies suggested that bio-markers (i.e., alpha, beta, and gamma) in specific brain areas could reveal the mental stress state [ 18 , 40 , 41 ].…”