부산대학교 도서관

Emotions have a significant impact on a person's behavior. Expression reflects how people perceive events, interactions, and judgment. It is possible to identify emotions and categorize them using several techniques, including electroencephalography (EEG) signals, gestures, facial expressions, speech patterns, etc. However, emotion recognition using physiological signals has gained popularity owing to its authenticity. In this study, 54 attributes were extracted from each EEG channel to gather emotional information. These attributes were then fed to multiple classifiers and the results were compared. The study attained a maximum accuracy of 89.52% for channel T7 - 24 using ensemble bagged tree classifiers.