23. DELT- Developing a Prediction Model for Academic Underachievement of International Students: The Case of K University in Korea

• In Event: Gallery Walk Online

Short Description

This study aims to establish a machine learning algorithm that predicts academic underachievement of international students. Log data compiled in the learning management system (LMS) was utilized in constructing the prediction algorithm. F1 score was used as the main evaluation index for performance comparison between models. Tuning process of hyper parameters followed, and finally an ROC curve was drawn in order to provide a visualization of prediction accuracy derived from the three algorithms.

Author

• Presenter: Hyo Jung Kim, Educational Technology, Department of Education, Korea University