Published Paper Details:

The high rate of student dropout in Massive Open Online Courses (MOOCs) is a serious issue with these courses. An effective MOOC student dropout prediction model can identify the reasons that cause students to drop out and provide insight into how to implement interventions to improve student success. For the prediction of student dropout in MOOC courses, many features and methodologies are available. The data from a self-paced math course, College Algebra and Problem Solving, given on the MOOC platform Open edX in collaboration with Arizona State University (ASU) from 2016 to 2020 are examined in this research. This research proposes a model for predicting student dropout from a MOOC course based on a set of variables derived from the daily learning progress of students.In the prediction, the Gradient Boosting Model technique from Machine Learning (ML) is performed, and validation factors such as accuracy, precision, recall, F1-score, Area Under the Curve (AUC), and Receiver Operating Characteristic (ROC) curve are used. With an accuracy of 87.5 percent, AUC of 94.5 percent, precision of 88 percent, recall of 87.5 percent, and F1-score of 87.5 percent, the model constructed can predict whether students will drop out or continue in the MOOC course on any given day. Shapely values were used to explain the contributing features and interactions for the model prediction