Student Performance Prediction and Optimal Course Selection: An MDP Approach

Abstract

Improving the performance of students is an important challenge for higher education institutions. At most European universities, duration and completion rate of degrees are highly varying and consulting services are offered to increase student achievement. Here, we propose a data analytics approach to determine optimal choices for the courses of the next term. We use machine learning techniques to predict the performance of a student in upcoming courses. These prediction form the transition probabilities of a Markov decision process (MDP) that describes the course of studies of a student. Using this model we plan to explore the effect of different strategies on student performance.

@inproceedings{backenkohlerStudentPerformancePredictionAnd2018,
 author = {Michael Backenköhler and Verena Wolf},
 booktitle = {Software Engineering and Formal Methods. International Conference on Software Engineering and Formal Methods (SEFM), Cham},
 doi = {10.1007/978-3-319-74781-1_3},
 editor = {Antonio Cerone and Marco Roveri},
 isbn = {9783319747811},
 pages = {40-47},
 publisher = {Springer International Publishing},
 title = {Student Performance Prediction and Optimal Course Selection: An MDP Approach},
 url = {https://link.springer.com/chapter/10.1007/978-3-319-74781-1_3},
 year = {2018}
}