Data-driven based student programming competition award prediction via machine learning models

Student performance prediction is one of the most important subjects for educational data mining. Generally, student performance prediction can be achieved by tracing the evolution of each student's knowledge states via a series of learning activities. However, previous studies mainly focus on predicting the grade of students. To the best of our knowledge, there are no studies focus on student competition award prediction, which is vital for students' development and learning process. In this paper, we aim to predict students' programming competition awards by building machine learning models with basic features and competition-related features extracted from 107 students. Specifically, we employ two kinds of data-driven-based features, which are basic features and programming competition-related features. The basic features are the personal and grade information of students in programming courses. The programming competition-related features are the performance data of students in the programming competition training activities. The empirical results show that the random forest model can achieve the best performance among all models with an Accuracy of 0.89, an F-measure of 0.88, a Precision of 0.90, and a Recall of 0.85. Moreover, the empirical results also show that the competition-related features are more influential features for predicting the students' competition awards, which should be paid attention to in the follow-up studies.