Mind the Gap: Confronting the Vast Divide Between CS Teaching and Machine Learning Pedagogy

The increasing demand for machine learning and deep learning (ML/DL) education spans K-12, college, and vocational training, highlighting the urgent need to prepare Computer Science (CS) teachers for these fields. This study aims to characterize the knowledge gaps and challenges experienced by CS teachers as they transition to teaching a rigorous ML curriculum. The paper attempts to achieve this in two distinct and self-contained ways: first, through a theoretical analysis conducted via a literature review examined through the lens of theoretical frameworks, and second, through an empirical qualitative analysis of a case study involving CS teachers transitioning to ML education. The empirical analysis echoes the findings of the theoretical analysis and sharpens them with additional insights. Our findings reveal significant difficulties, such as relearning mathematical foundations, adapting to new problem-solving paradigms, and developing ML-specific pedagogical content knowledge. Notably, the existing expertise of experienced CS teachers has limited relevance to ML/DL education, raising the question of why we focus mainly on CS teachers as the potential teaching workforce to train. The discussion integrates the theoretical and empirical findings to offer conclusions and recommendations for educational institutions, policymakers, and teacher training programs in enhancing ML/DL teaching capacities across various academic levels.