A Deep Learning-based System for Detection At-Risk Students

The digital revolution has an impact on educational systems, which makes a significant shift from traditional education to e-learning. Nowadays, many universities throughout the world use e-learning platforms as part of their learning approach. One of such systems is Massive Open Online Courses (MOOCs), which has seen great success and an increase in the number of students who enrolled in such learning vision. However, this method of learning suffers from the problem of previous students'dropouts. In this paper, LSTM, GRU and BiLSTM are usedas deep learning techniques to develop an intelligent system that is able to detect at-risk students at early stages. We ran experiments using a Harvard dataset to assess the performance of the proposed method.