Demystifying AI in Mathematics Education

Zusammenfassung

AI and machine learning (ML) rely fundamentally on mathematical modeling. Interestingly, many of the core mathematical techniques underpinning ML are quite elementary – often accessible with high-school level mathematics. This observation suggests that AI education should not be limited to computer science courses but should also be meaningfully integrated into mathematics curricula. In my talk, I will explore various strategies for simplifying and teaching key mathematical ideas behind different ML methods (e.g., Support Vector Machines). These methods are deeply rooted in mathematical concepts from linear algebra and calculus – areas of mathematics that students often find abstract or unengaging. However, the mathematical methods used are often elementary and accessible with high-school knowledge, such as analytical geometry. By addressing real-world, data-driven problems from AI within the context of mathematics education, we have a unique opportunity to make mathematical concepts more relevant and exciting for students, while also fostering a deeper understanding of AI, including its risks and potentials. I will present tested digital learning materials for high-school students and will critically examine which underlying mathematical concepts can be explored in the classroom (‘white-box’), and which may need to be treated as ‘black-box’ due to their inherent complexity. The material was developed within the CAMMP project (www.cammp.online) and focuses on the problem-oriented development of the mathematical foundations of AI methods. I aim to show that it is indeed possible to embed AI into mathematics education in a meaningful way.

Sarah Schönbrodt

Assistenzprofessorin @ Universität Salzburg

Forschung im Bereich Mathematikdidaktik und KI-Bildung