In honor of the Department of Statistics' 50th Anniversary, we are highlighting our outstanding faculty and alumni of the past half-century.

Michelle Dunn

Michelle Dunn (Ph.D. 05) first learned about statistics on the 1,300-mile drive to college, from her hometown of Memphis, Tennessee. All Harvard freshmen had to pass a Quantitative Reasoning (statistics) exam upon arrival, or else were required to take an introductory statistics class. Although Michelle passed the exam, she appreciated the usefulness of statistics so much that she enrolled in multiple statistics classes over the next four years as part of her applied mathematics major. When she decided to continue her studies in statistics with a master’s degree, she conducted a non-scientific survey of Harvard graduate students, asking their advice about where she should go next.  Knowing she was Bayesian at heart, they unanimously encouraged her to go to Carnegie Mellon.    It was the only school to which she applied. “I first met Michelle shortly after she arrived when she was assigned to me as a TA,” Prof. Jay Kadane recalled.  “She told me she would be happy to work as my TA under one condition: that I would take academic integrity violations seriously. (Apparently this had been a serious issue at her undergraduate institution). I so promised, and she did a swell job as a TA. (It also started me on a quest to be much more cognizant of the matter than I had been before),” he said.  After graduating with a master’s degree, Michelle worked as a computational statistician, honing her skills in programming and algorithm development.  After four years, she returned to Carnegie Mellon as a PhD statistics student.
 “We stayed in touch during those years and, when she came back, she was very concerned about improving education, and I had some contacts that led to data from the Pittsburgh school system,” Jay said.  “This became her ADA (Advanced Data Analysis) project, and resulted in a published paper. For her thesis, she decided she wanted to work on cybersecurity, and finished a thesis on that subject,” he said. Working with Jay, Michelle’s projects included applying MCMC and particle filtering, first to education, and later to detecting anomalies in Internet traffic.  During this time, she and her husband, Steve, became parents to identical twin girls.  After graduation, Michelle focused on raising the twins until they started school.  She then joined the National Cancer Institute as a Program Director, managing a portfolio of statistical methods grants.  In this position, she was afforded the opportunity to learn about the exciting, cutting-edge research of statisticians from across the country.
While at the NCI, Michelle volunteered to be part of the team developing the Big Data to Knowledge Initiative. Combining her interests in education, statistics, and computation, Michelle led the development of the training component of the BD2K Initiative, which invests about $15 million a year in data science education and workforce development.   “Michelle has been a very important voice at NIH on behalf of statistics, shaping the role of statistics in funding initiatives especially as Big Data has become prominent,” Prof. Robert Kass said.  “It is great that she's there, and I'm proud to count her as an alum,” he said.  The NIH Associate Director for Data Science, Phil Bourne, invited Michelle to join his team, making her the only statistician in the office.  While in this position as Senior Advisor for Data Science, Michelle collaborated with the National Science Foundation to build the joint NSF-NIH QuBBD (Quantitative Biomedical Big Data) program, continued to expand the BD2K training program, and organized a portfolio of training courses for the NIH workforce.  Being unencumbered by tradition or extensive knowledge of NIH, Michelle was able to develop creative new programs, including a portfolio of Open Educational Resources and an Educational Resource Discovery Index (ERuDIte), for finding resources to learn about data science.  She remains active in the statistical community, serving as chair of the American Statistical Association’s (ASA’s) Committee on Funded Research, a member of ENAR’s advisory boards (RAB and RECOM), and a member of a National Academies of Science Roundtable.  She also delivers presentations at ASA board meetings and chair workshops, at professional society meetings, and at universities across the country.  “Michelle throws herself into whatever her top priority is at the moment, and does excellent work on it. I can't wait to see what's next,” Jay said.