Crafting an Undergraduate Course at the Intersection of Machine Learning and Software Engineering

Abstract

This is a work-in-progress innovative practice short paper. It describes exploratory efforts to craft a new undergraduate course at the juncture of Machine Learning (ML) and Software Engineering (SE). Our motivation comes from recent industry practices and challenges to adopt and deploy ML as part of software development projects.We leverage reported industry experiences by ML practitioners, research findings from academia, as well as student feedback, in order to create a new undergraduate course where ML and SE cross paths. Toward this end, we conducted an empirical study -- student surveys and interviews of ML professionals. Questionnaires helped us measure students’ interest in learning how to leverage SE topics to develop ML-based systems, and to gauge their expectations and concerns about the course. Our interviews with ML professionals helped us solicit and collect recommendations about the content of our course.In summary, our paper presents qualitative results of an empirical study launched to help us craft a new course on ML and SE. We discuss a list of reported challenges from ML professionals and explain how we match them to related academic learning objectives and course content. We also describe how student feedback including their expectations and concerns helped us in this effort. We hope that such a course will not only prepare our students for the workforce but will provide a promising example for other institutions to follow.