Data science is inherently multi-disciplinary, requiring training in statistics, computer science, and one or more domain areas. While no one person may have all the requisite skills, as a community of data scientists, we need to encompass these broad skills and expertise as a set with sufficient cross-disciplinary fluency to enable us to communicate and collaborate effectively. Traditional university curricula, however, do not provide this kind of broad, cross-disciplinary data science training.
The goal of this working group is to learn from past training and education activities at our partner (and other) institutions to develop a new data science curricula that appeal to our heterogeneous data science community. We hope to overcome siloed approaches that waste resources and discourage interdisciplinary learning and discovery by developing instruction that is flexible yet keeps sight of a common core of skills, knowledge, and language.
Working with our partners at New York University and the University of Washington, we plan to develop effective bootcamp materials (covering many areas and skill levels) and will build on existing successful bootcamps, such as Software Carpentry, to extend their coverage to other data science topics. Additionally, we will offer tune-up camps and continuing-education activities. We also plan to develop several online courses in a variety of areas and for different skill levels as well as host hackathons and seminars.