Feature Selection and Comparison for the Emotion Recognition According to Music Listening

Recently, researches on analyzing relationship between the state of emotion and musical stimuli using EEG are increasing. These research shows that a selection of feature vectors is very important for the performance of EEG pattern classifiers. In this paper, we apply feature extraction methods, which were reviewed in the previous, to DEAP data for the emotion recognition. We limit to analysis features in time-domain for this research. To evaluate the feature vectors, the Relief algorithm and the Bhattacharyya distance are used. According to result, the power of signal is better for the emotion recognition than the other feature.