CPH Tech Policy Brief #5

Algorithmic admissions: Predicting college dropout risks with granular academic data

This fifth edition of the CPH Tech Policy Brief presents insights from a new research study by Magnus Lindgaard Nielsen, Jonas Skjold Raaschou-Pedersen, Emil Chrisander, Julien Grenet, Anna Rogers, David Dreyer Lassen, and Andreas Bjerre-Nielsen: “Transforming Prediction Policy: How Novel Machine Learning Methods Can Improve Higher Education Admission.”

The development of artificial intelligence and machine learning in the recent decade has been nothing short of meteoric. This CPH Tech Policy Brief investigates how the recent advances in AI and machine learning could be employed for algorithmic policy making.

We summarize our recent work that outlines how to employ the recent advances in machine learning and that provides evidence of how such AI-based systems can improve policy decisions while preserving the same level of fairness as human decisions.

The context of the study is admissions to higher education in Denmark where we find that the new methods outperform standard approaches and are more objective in assessing academic aptitude than current admission criteria when comparing different students’ backgrounds.