The Role of Extreme Learning Machine in Multi Classification based Emotion Recognition using EEG Signals

Emotion is one of the most significant parts of the daily interaction among people. Electroencephalography (EEG), which provides the transfer of physiological signals of the nervous system to the computer environment, plays an important role in emotion recognition. In this study, the performance of Extreme Learning Machines (ELM) in multiple classification based emotion recognition on DEAP data set is examined. Features required for the classification are obtained from time domain and frequency domain. Emotions are classified as binary and multiple according to their arousal and valence values, and the classification results are also compared with Support Vector Machines (SVM).