Human Emotion Variation Analysis Based On Eeg Signal And Poms Scale

Youjun Li,Haiyan Zhou,Jianhui Chen,Jiajin Huang,Meng Chen,Yan Liu,Ning Zhong
DOI: https://doi.org/10.1007/978-3-319-47103-7_8
2016-01-01
Abstract:Emotion is considered as a critical aspect of human brain behavior. In this paper, we investigate human normal emotion variation for a long period without stimuli. Eight subjects participated in the experiment for seven days. The EEG signal and POMS scale of the subjects were collected in the experiment. After data collection and preprocessing, Pearson correlation analysis and multiple linear regression analysis were carried out between EEG features and POMS emotion components. The results of Pearson correlation analysis show that the correlation coefficient of EEG features and POMS emotion component range from 0.367 to 0.610 at 0.01 significant levels. Based on this, multiple linear regression models are built between POMS emotion components and EEG features. With these models, the POMS scales of the subjects can be predicted such that the R-2 between the prediction scale and real scale ranges from 0.329 to 0.772; the emotion of 'Depression-Dejection' has the lowest R-2 (0.329); and the 'Negative Emotion' has the highest R-2 (0.772).
What problem does this paper attempt to address?