Investigating Sex Differences in Classification of Five Emotions from EEG and Eye Movement Signals

People generally agree that emotion processing differs between male and female. However, current hypothesis of sex differences needs more objective evidence and quantitative assessment. In this paper, we investigate the sex difference in classifying five emotions from eletroencephalograph and eye movement signals. We adopt two neural-network-based classifiers to objectively investigate sex differences from different perspectives. From experimental results, we find the following three observations: (1) a general higher accuracy of same-sex strategy suggests sex-specific factors have influence on emotion classification; (2) both blink duration and frequency differ from female to male and they are negatively correlated under different emotional states; and (3) there are larger differences of brain activities in the Theta, Alpha, and Beta bands between male and female for disgust, sad, and neutral emotions.