Classification of Emotional States and Familiarity based on EEG Signals

Emotional reactions and familiarity towards the multimedia stimuli are closely connected and have a strong effect during tasks related to the classification of emotion and cognitive activity. The current study presents spectral activity-based EEG features for simultaneous classification of emotional...

Full description

Saved in:
Bibliographic Details
Published in2021 International Conference on Biomedical Innovations and Applications (BIA) Vol. 1; pp. 54 - 57
Main Author Feradov, Firgan
Format Conference Proceeding
LanguageEnglish
Published IEEE 02.06.2022
Subjects
Affective computing
Classification algorithms
EEG features
Electroencephalography
emotion classification
familiarity classification
Multimedia databases
Support vector machines
Task analysis
Technological innovation
Online AccessGet full text

Cover

More Information
Summary:Emotional reactions and familiarity towards the multimedia stimuli are closely connected and have a strong effect during tasks related to the classification of emotion and cognitive activity. The current study presents spectral activity-based EEG features for simultaneous classification of emotional states and level of familiarity towards multimedia stimuli. The experimental evaluation of the examined features was conducted using the SEED database and three classification algorithms - polynomial SVM, kNN and C4.5 - that were employed in a 9 class classification task. A maximum mean classification accuracy of 95.2% is reported.
DOI:10.1109/BIA52594.2022.9831267