A New Emotional Classification Method Based on the Combined Characteristics of EEG Signals
Download as PDF
DOI: 10.23977/csic.2018.0934
Author(s)
Tengfei Wang, Guangyuan Liu
Corresponding Author
Guangyuan Liu
ABSTRACT
The electroencephalogram (EEG) signals has been widely used in emotion recognition. The entropy features are often used in the emotional recognition using EEG. However, its recognition accuracy remains to be improved. In this study, a new EEG feature combining the frontal asymmetry and differential entropy features is used to classify negative and positive emotion. The result of our research shows that the combined feature is better than the entropy features in emotion recognition. The average classification accuracies of the frontal asymmetry combined with differential entropy features and entropy feature in our study are 72.1% and 67.7% respectively. This result indicated that this combination feature is more suited for emotional classification.
KEYWORDS
Eeg Signals, A New Emotional, Emotional Classification