���ʵ���

Why is machine learning difficult? Triangulating novice challenges and common errors

Lisa Zhang
University of Toronto

Abstract: Machine learning (ML) is often seen as a challenging area for undergraduates. This talk examines two key sources of difficulty: (1) applying ML in practice, and (2) understanding the mathematics behind ML. Drawing on two recent studies, I will share findings on common errors students made in ML projects at the University of Toronto [1], as well as how students themselves describe their interests and challenges [2]. This latter study produced specific hypotheses as to why math in ML can be difficult. Triangulating these results with related work in the computing education literature provides a more complete picture of the sources of difficulty in ML learning and suggests how ML educators can design more supportive learning environments.

[1] Zimmermann et al. “Common Errors in Machine Learning Projects: A Second Look”  

[2] Sibia et al. “Student Perspectives on the Challenges in Machine Learning”  

Bio: Professor Lisa Zhang is an Assistant Professor, Teaching Stream, in the Department of Mathematical & Computational Sciences at the University of Toronto Mississauga. Since 2018, she has taught a variety of undergraduate machine learning courses and contributed to research in machine learning education. She was named an EAAI New and Future AI Educator in 2020 and has contributed five peer-reviewed assignments to the EAAI Model AI Assignment repository. More recently, she received the UTM Teaching Excellence Award (2024) and the University of Toronto Cheryl Regehr Early Career Teaching Award (2025). Her teaching and research interest also includes writing in computer science: she led the launch of the Computer Science Writing Development Initiative at UTM, and  an experience report on this work received a Best Paper award at SIGCSE, the largest international conference in computing education.

This guest talk is hosted by Assistant Professor Juho Leinonen, Department of Computer Science.