Most people use spring break as a time to unwind, but it was business as usual for Dick De Veaux, professor of statistics. During his two-week break from teaching, he went on a total of 10 plane rides to give talks on his work, lead a forum on teaching, do consulting work and pay a visit to his son in college. “That’s not atypical of my schedule,” De Veaux said.
For his impressive work in the field of applied statistics, De Veaux was recently named the Mosteller Statistician of the Year by the Boston Chapter of the American Statistical Association.
The award, which is named for the renowned statistician Charles Frederick Mosteller, recognizes De Veaux’s influential research on data mining methodology and its applications. De Veaux extends his expertise in data analysis to address problems in science and industry. “I apply things that I read in the statistical literature to other fields,” he said.
To describe his work in broad terms, De Veaux used the example of data mining by credit card companies that hold large data sets. “They want to find information that is in there, the data, like needles in haystacks,” he said. “They are looking at the way you are using your credit cards, and thinking about what offers they can extend to you to maximize your spending.” His task is to solve such problems using statistical models.
De Veaux has consulted with more than 30 Fortune 500 companies, including Hewlett-Packard, Alcoa, American Express, Bank One, DuPont, Pillsbury and General Electric. However, he made clear that he values teaching as much as his other professional endeavors, and praised the College for its exceptional support for teaching. “In large universities, you get no support [for teaching]. At those places research is 100 percent of what you do,” he said.
Having previously taught graduate engineering students at Princeton, De Veaux admitted that he prefers working with undergraduates, including those who are less inclined to take courses offered in the mathematics and statistics department. “The only student who is frustrating is the student who just doesn’t want to get engaged,” De Veaux said. “Really, most professors love teaching anyone who wants to learn.”
Several of De Veaux’s students spoke with enthusiasm about their experience in his class. “I feel really lucky to be taking his tutorial because he’s such a leader in the field,” said Jon Prigoff ’08, who is taking De Veaux’s tutorial on, naturally, data mining. “He’s such a great resource and through him we have access to some really powerful tools, but he really keeps the focus on us figuring everything out for ourselves.”
De Veaux is a recipient of numerous accolades in addition to this most recent award. He particularly prizes his Lifetime Achievement Award for Exceptional Dedication and Excellence in Teaching, presented by the Engineering Council at Princeton, “because it was totally student generated,” he said.
In addition, he is a recipient of the Shewell Award for Best Paper and Presentation and the Frank Wilcoxon Prize for Best Practical Application Paper from the American Society of Quality Control. He was also elected fellow of the American Statistical Association and chairman of the Gordon Research Conference.
De Veaux earned his A.B. in mathematics and his B.S.E. in civil engineering from Princeton and his Ph.D. in statistics from Stanford. Before joining the College’s faculty in 1994, he taught in the department of civil engineering and operations research at Princeton and in the department of statistics at the Wharton School at the University of Pennsylvania. Since coming to Williamstown, he has also been a visiting research associate at the Institut National de la Recherche Agronomique in France and a visiting professor at Princeton.
He has co-authored three critically acclaimed statistics textbooks that are used in classrooms throughout the nation. His works are noted for using an engaging writing style that appeals to students from varying academic disciplines. “The text incorporates compelling examples derived from the authors’ wealth of consulting experience and encourages readers to learn how to reason with data,” said an online review of Stats: Data and Models. De Veaux is currently involved in writing a fourth textbook.
Additionally, he plans to write a book on data mining at some point. “I’m not sure whether it will be a popular text … or a textbook,” he said.
De Veaux is also a standout among statisticians for a more unconventional reason: few, if any, statistics professors have once been professional dancers. De Veaux holds a master’s degree from Stanford in dance education, and he regularly teaches modern dance at the College as a Winter Study course.
Despite his international reputation, De Veaux remains humble and does not take himself too seriously. “When you get to a certain age, people just start giving you awards,” he said. “I think there is some truth to that.”