A dropout prediction method based on time series model in MOOCs

Abstract In recent years, MOOCs has enjoyed great popularity due to its convenience and openness. However, with the development of MOOCs, high dropout rate aroused extensive attention. By analyzing data students’ behavior then predicting whether students are at risk dropout, it can improve course completion rate. Most existing methods relying on feature engineering sequential characteristic is not effectively utilized. this paper, we propose a time series model named CNN-LSTM-ATT, which focuses more local valid information temporal data. Through experiments public dataset, shows that proposed predict behavior.