Youth Learning Machine Learning through Building Models of Athletic Moves

Abstract

Machine Learning-based (ML) technologies impact many facets of our lives. Given ML's ubiquity, and the ways it offers creative computational possibilities distinct from programming, we believe it could be a powerful tool for youth to leverage in making, creativity, and play. We investigate how youth with no programming experience can incorporate ML classifiers into athletic practice by building models of their own physical activity. In this paper, we describe a design experiment exploring how to introduce youth to making ML models within the context of their athletic interests. We present AlpacaML, an iOS application that connects to wearable sensors and allows young people to model physical movement using an ML classifier, and detail its use in a three-hour workshop with middle- and high-school athletes. We found the youth were able to collect data, build models, test and evaluate models, and quickly iterate on this process. We finish with a discussion of why this is a promising direction for the incorporation of Machine Learning into novice youth making, exploration, and play.