Data Science ARES: Rebecca Barter
Join us at the Data Science Applied Research and Education Seminar (ARES) with:
Dr. Rebecca Barter
Department of Statistics
University of California, Berkeley
Teaching Data Science in the Real World
Data science is often viewed as a subfield of statistics and/or computer science, and is thus often taught accordingly with a focus on mathematical and computational theory, rather than a focus on the real world. As universities are starting to dip their toe into the new field of data science, they are struggling to adapt their age-old statistical- and computer science-based teaching approaches to the new data science era. It is not enough to simply update old classes with new topics, such as deep learning and causal inference (A/B testing). There is a strong need for a new approach to teaching that connects mathematical concepts and symbols to the real-world in the context of domain problems, and incorporates empirical data skills, such as asking data-driven questions; dealing with real, messy data; and scrutinizing data-driven results. In this talk, I will discuss my own experience making my way through a traditional statistics education, and introduce a book that I am co-authoring with Professor Bin Yu (Member, US National Academy of Science) called “Veridical Data Science: The Practice of Responsible Data Analysis and Decision Making”, based on Professor Yu’s applied statistics class (STAT 215A) taught at UC Berkeley. Veridical data science is built around the three guiding principles of data science: Predictability, Computability and Stability (PCS), that unifies and expands on ideas and best-practices from both statistics and machine learning. Our book aims to bring this new way of practicing and teaching data science to a wider audience.
Rebecca completed her PhD in Statistics from UC Berkeley advised by Professor Bin Yu in December 2019. During her PhD, she authored the superheat R package, maintained a popular blog, and embarked on a range of data science projects primarily focusing on applications of machine learning and causal inference to problems in healthcare. Now as a postdoc, she is co-authoring a book with her PhD advisor and mentor, Professor Bin Yu, called Veridical Data Science: The Practice of Responsible Data Analysis and Decision Making, which teaches data science using intuition and critical thinking from a real-world perspective.