Uncovering of learner’s confusion using EEG Data during online learning

Confusion is a very common phenomenon during learning. Sometimes learners are not able to grasp the actual idea of whatever they are being taught. Confusion is a symptom that makes you feel that you cannot think clearly. The proposed work is based on the detection of the confused mental state of students during online learning. For this purpose, a publicly available dataset has been used. After truncating the data points to the same size, Bidirectional LSTM(Long Short Term Memory) network has been used to classify, whether the student was confused or not. Using the proposed method, 75% accuracy was achieved. We have also analyzed the effect of different epoch sizes and size of validation split on the accuracy and it has been found that the highest accuracy was achieved on epoch size 80 and validation split 80:20. The proposed work is able to detect learners' confusion using EEG data what will help the course designer to design the course accordingly.