“…There are, however, two approaches to the classification of emotion through EEG and these include the machine learning and neural network approaches. a) Machine learning approach: Some of the methods usually applied include decision tree (DT) [71], naïve bayes (NB) [72], quadratic discriminant analysis (QDA) [73], k-nearest neighbors (kNN) [58], [74], [75], linear discriminant analysis (LDA) [14], relevance vector machines (RVM) [67], xtreme gradient boosting (XGBoost) [76], support vector machine (SVM) [77]- [79], AdaBoost [80], logistic regression via variable splitting and augmented lagrangian (LORSAL) [81], random forest (RF) [56], [82], and graph regularized extreme learning machine (GELM) [83]. b) Neural network approach: This method include artificial neural network (ANN) [61], [63], [84] deep belief networks [70], [85], convolutional neural network (CNN) [40], [46], [86], [87], long short-term memory (LSTM) [66], generative adversarial networks (GAN) [88], capsule network (CapsNet) [45], [62], and hybrid methods [4], [44], [69].…”