As part of the NSF Data Science Seminar Series, Michael I. Jordan—chair of the University of California, Berkeley Department of Statistics—made his case for why one should think of data science as the combination of computational and inferential thinking.

