2026 ASEE Annual Conference & Exposition

In this GIFTS paper, we present an introductory module that builds learners’ holistic understanding of Machine Learning (ML) alongside the associated data pipeline by transforming the classroom and its occupants into an ML model. Heavy math-based introductions to Machine Learning are the norm in university settings, which can be overwhelming and obstructive to learners who are interested in getting into the field, especially early university undergraduates. The traditional method of teaching Machine Learning leaves students, regardless of grade level, overwhelmed and unable to intuitively internalize the machine learning pipeline. Here, we present an activity that provides mechanical sympathy for how machines actually learn, one that requires no prerequisite knowledge and is appropriate for learners from middle school to university settings. Better yet, our activity is computer-independent; it disrupts students’ comfortable familiarity with computers as a black box solution, allowing learners the freedom to think about computers differently. By removing the computer from the equation, learners become immersed in the activity and are more collaborative and communicative as a result.

In Becoming an ML Model, students become the machine and play a guessing game to ‘learn’ or recognize patterns. The different stages of the activity mirror the ML train-test-deploy pipeline to provide participants an intuition into the ML workflow, while simultaneously introducing important concepts such as bias in data, accuracy, model pattern recognition, as well as key limitations of ML. Furthermore, the activity serves as an icebreaker for students as it encourages an excited and high-stakes conversation among all students on the first day. The evaluation of Becoming an ML Model comes from a post-activity recap discussion, concurrent instructor observations, and pre-post student-notebook reflections. Based on these methods, the activity has proven a success in achieving and, in some cases, surpassing its goals: during the activity, most, if not all, students were highly engaged over the duration, to the extent of having to quiet the class down after each discussion due to the excited deliberation over the activity prompt. Not only did learners correctly answer the recap questions with confidence, they also engaged in meaningful classroom discussion to understand and unpack some of the more philosophical issues that came up. After class, learners often stay behind to tell us how useful, helpful, and engaging the activity was.