Using Machine Learning to Predict Mathematical Performance

Abstract

In empirical educational research, it is of great interest to predict student performance. In contrast to other disciplines, however, machine learning methods for identifying promising predictors are not yet widely used. We will use a machine learning approach to study the effect of learning strategies and cooperative behaviors of German mathematics students with respect to exam grades. A small outlook will be given on how machine learning methods can be integrated into the education of young researchers in empirical educational research.