EEG classification with limited data: A deep clustering approach

Mohsen Tabejamaat, Hoda Mohammadzade, Farhood Negin, François Brémond

Abstract: Highlights•A novel training strategy is proposed to mitigate the overfitting of deep neural networks.•A clustering approach is introduced to address outlier issues in EEG data.•A novel data augmentation method is proposed for EEG signals.