In this study, we conducted EEG-based emotion recognition on arousal-valence emotion model. We collected our own EEG data with mobile EEG device Emotiv Epoc+ 14 channel by applying visual-aural stimulus. After collection we performed information measurement techniques, statistical methods and time-frequency attributes to obtain key features and created feature space. We wanted to observe the effect of features thus, we performed Sequential Forward Selection algorithm to reduce the feature space and compared the performance of accuracies for both all features and diminished features. In the last part, we applied QSVM (Quadratic Support Vector Machines) to classify the features and contrasted the accuracies. We observed that diminishing the feature space increased our average performance accuracy for arousal-valence dimension from 55% to 65%.
emotion estimation feature selection support vector machines EEG
Birincil Dil | İngilizce |
---|---|
Konular | Bilgisayar Yazılımı |
Bölüm | Araştırma Makalesi |
Yazarlar | |
Yayımlanma Tarihi | 1 Aralık 2019 |
Gönderilme Tarihi | 24 Aralık 2018 |
Kabul Tarihi | 27 Haziran 2019 |
Yayımlandığı Sayı | Yıl 2019 Cilt: 23 Sayı: 6 |
Sakarya University Journal of Science (SAUJS)