Student Competitions

Duke Student Team Wins Annual DataFest

• By Michael Hart
• 04/29/15

Four students from Duke University spent a weekend recently analyzing the behavior and preferences of automobile buyers. The result of their 48-hour data analytics marathon was an award from the American Statistics Association's DataFest for Best Visualization of Data.

The Duke University team was one of more than 25 from colleges and universities that participated in the annual competitive data analytics event held each year since 2011. More than 100 college students gathered at seven campuses around the country to pore over data provided by car shopping website Edmunds.com.

They were asked to select data that was available, analyze it and then present it to a panel of judges in a presentation with a time limit, using only two PowerPoint slides. All the teams were assisted by roving consultants who were graduate students, faculty members and industry professionals.

The four Duke University team members — David Clancy, Tori Hall, Michael Lin and Gregory Poore, all scheduled to graduate in 2016 — participated as the Bayes' Anatomy team, focusing on analyzing and visualizing auto consumer preferences and behaviors.

"The students found a lot of unexpected and interesting stories to tell using the car shopping data from Edmunds," said Duke University Assistant Professor Mine Cetinkaya-Rundel, one of the founders of the competition. "This made them appreciate the importance of data analysis skills and made them excited for working with such data in the future."