News

DataFest growth continues; record number register for event


April 2018

Jay Murdock

With a name like “DataFest”, it’s not surprising that to see the event’s growth in popularity, you only have to look at the numbers.

In the event’s second year, 88 students participated. This year, 173 have registered.

“To have 173 people that want to dedicate an entire weekend of their college experience to this, to have schools calling us to see if they can participate, it just means there’s a lot of attention on this, that the students understand the value of learning these skills and that the Center for Analytics and Data Science (CADS) is delivering value back to the campus by providing a place to bring people together in the name of data and data science,” CADS director Dr. Allison Jones-Farmer noted.

A cooperative effort between the College of Arts and Science, the College of Engineering and Computing, and the Farmer School of Business, DataFest is designed to give students the opportunity to work with real-world data. Teams of 3 to 5 students use the data to find answers - so long as they can find the questions first.

“We give students the opportunity to work very independently on a pretty big data problem with some high-level questions involved that can be solved and attacked many different ways,” assistant professor of statistics Dr. Thomas Fisher said.

“It gives students an opportunity to see the kind of data that they don’t get to see very often, and they have to wrestle with what real-life data means - the errors, the clumsiness of it that is oftentimes scrubbed for them before they go into class. It forces them to realize what goes into it,” CADS assistant director Lindsey Holden pointed out.

After receiving the data Friday evening, the teams work in FSB through the weekend. They have until Sunday morning to sift through the data, figure out what it means, and find meaning in it.

“On Sunday morning, they will make a presentation. They get six minutes to tell us something important in what they found,” Fisher said.

The top teams from each of five morning presentation sessions move on to the afternoon presentation, which take place before a group of judges comprised of industry partners. “You have to convince other people that what you did is important and meaningful,” Fisher explained.

“We try to treat it like a developmental environment so that nobody feels bad if they don’t have the same skill set as one of the winning teams. We hope that everybody walks away knowing something more than they came in knowing,” Farmer-Jones said.

The top three teams win cash prizes - up to $1,000 for the winning group.

The event isn’t just for data scientists. Fisher said any student can participate - and should.

“I would recommend any student try it at least once. I say ‘once’ because it’s a hard weekend. You’re working 40 hours, basically, in two and a half days,” he remarked.

But despite the difficulties, a sizable number of this year’s registrants have participated in one of both of the previous DataFests.

“It says that we do a good job and they find value in it,” Farmer-Jones said. “Every data set is different, you learn different things every year, and it’s an opportunity for them to grow and to mature.”

“It’s something we’re really proud of,” she said.

The event is sponsored by Procter & Gamble and Vantiv (now Worldpay), along with CADS partners.

DataFest participants having discussion in work area