A Comparative Prediction of Dropouts of Students from Massive Open Online Courses Using Machine Learning Algorithms

Institutes and online platforms must detect dropout rates in MOOCs because early discovery enables proactive strategies to make course completion and learning more efficient. We propose an approach to predict student dropouts in MOOCs through machine learning algorithms. The methodology begins by selecting features through techniques that find good predictors of student attrition. Many features are ranked, including demographics, history of past academic performance, course engagement metrics, and social interactions with peers. We use ensemble learning to develop strong predictive models such as decision trees, logistic regression, and support vector machines to achieve the strengths of individual components and minimize their weaknesses. These models are then trained on selected features using various machine learning algorithms. Evaluating results We use large-scale MOOC enrollment and activity data.The research is compared with individual classifiers and several state-of-the-art techniques of machine learning algorithms and after that it is said that ensemble models do indeed show better predictive power than all the other methods used.