Predicting MOOC Dropout over Weeks Using Machine Learning Methods

摘要: With high dropout rates as observed in many current larger-scale online courses, mechanisms that are able to predict student become increasingly important. While this problem is partially solved for students active forums, not yet the case more general population. In paper, we present an approach works on click-stream data. Among other features, machine learning algorithm takes weekly history of data into account and thus notice changes behavior over time. later phases a course (i.e., once such available), significantly better than baseline methods.