ABSTRACT Name : Mohamad Dwiyan Rahmanianto Program : Computer Science Title : Student at Risk Detection in Online Programming Environment Difficulties in learning programming on the introductory programming course driving research in learning analytics to understand the learning process in this course. This study build prediction model for identifying struggling students in learning introductory programming by using student programming activity data from http://exercise.ui.ac.id/ web platform. The algorithm used is decision tree, support vector machine, K-nearest neighbor, dan logistic regression. Students who predicted strugling can be intervened to help them with the study. Keywords: prediction, Classification, learning analytics, programming