Estimation of Context-dependent Feature based on CNN-CAM

This study aims to design EEG meta-BCI system that estimates a learning strategy of the subject during?performing two-stage MDP. This model utilizes context data extracted by arbitration model to train CNN-CAM deep learning model. Well trained meta-BCI is able to estimate not only subject’s learning strategy but also visualize context dependent EEG features.

Contact: Dongjae Kim (kim10481@kaist.ac.kr)