In 2011, Rob Gould, vice-chair of undergraduate studies at UCLA’s department of statistics, created the DataFest, a team competition in which students were given 48 hours to analyze a complex dataset and provide intelligent, creative insight before a panel of judges. The competitive nature of the event was designed to rejuvenate students’ academic performance while giving them a taste of real-world, current data. Since that first event, DataFest has grown significantly and is now hosted by five major schools nationwide.

“I’m really excited and pleased to see how DataFest has grown. In 2011, we had about 30 students at UCLA, and now we had almost 400 students at 15 colleges and universities around the country, all analyzing the same dataset,” said Professor Gould. “I think this speaks to a real hunger that students feel for being provided with challenging problems in a fun and supportive environment.”

Each year, the data provider is kept secret until the beginning of each event. GridPoint provided “secret” dataset this year and students were challenged not only to glean imaginative analysis, but were also required to learn about the consumption and reduction of energy in a short amount of time. Data-crazed students participated in DataFests that were hosted at Duke University, Princeton University, Emory University, UCLA and the Five Colleges.

Kyle McCarter, director of solutions engineering at GridPoint, who received his B.S.E. from Duke University in 2005, said attending the event at his alma mater gave him a new perspective on the data GridPoint deals with daily. “We deal with difficult data. Time-series data was new to a lot of people, even the graduate students. Additionally, there is a lot of contextual information (like how an HVAC system works) that is necessary to model things properly,” he noted.

Overall, Kyle reported that students’ approach to GridPoint’s data was compelling and found that their analysis offered valuable ideas for future exploration.

Congratulations to this year’s winners and thank you to the DataFest organizers for featuring GridPoint! 

Duke University:
Best in Show: “Wolfram Alphas”: Mao Hu, Avinash Moondra, Sankash Shankar, Daniel Zhang, Zenia Zheng, Josh Latner, Meghan Scanlon; “D 4 C”: Alvin Kang, Paul Kim, Lilly Park, Kevin Wu, Eric Wu
Best Use of External Data: “Team Clairvoyant”: Ashish Agarwal, Yu Fan, Fangjian Guo, Yizhe Zhang, Weiqiang Zhu; “The Fantastic 5+1”: Xirui Diao, Mingqiang Fu, Xiangcheng Gao, Yi Rong Hoo, Ruisi Zhan, Brianna Bai; “Data Cruncherz”: Travis Britain, Brittany Cohen, Varun Himamshu, Gabrielle Levac, Heather Shapiro, Philip Srevrev
Best Visualization: “Cougar Bait”: Lorin Crawford, Nghi Nguyen, Derek Owens-Oas, Shinichiro Shirota, Xi Chen, Joseph Futoma, Christoph Hellmayr, Matt Johnson, Victor Pena; “Spectral”: Bryan Davis, Fuhui Fang, Yinan Fang, Jie Gao

Princeton University:
Best in Show: “DJ Hydrazine”: Harold Li, Nathaniel May, David Zhao
Best Visualization: “The Outliers”: Ujjwal Dahuja, Mihaela Curmei, Arunraj Balaji, Mark Aksen, Lydia Liu, Kevin Lin
Best Use of External Data: “WuTang Financial”: Farhan Abrol, John Anderson, David Jensen, Leor Klainerman, Robert Sami

Five Colleges: Smith College’s “The P-Valuables”: Deidre Fitzpatrick Michele Handy, Maja Milosavljevic , Sara Stoudt, Dana Udwin

Emory University:

Best Insight and Best Visualization: “Data Miners”: Kirk Hadley, Aiming Nie, Michael Zhai, Tony Wang, Fei Gao; “ANOVA One Bites the Dust”: Kyle Niezgoda

UCLA:

Best Insight: “Oops I Dataed Again”: Siena Duplan, Kayla Frisoli, Kimberly Mauleon, Greta Olesen, Daniel Smith; “SC Statistics”: Chris Funaki, Sri Ramesh, Bradley Rava, Tsz Chan, Matthew Norris
Best Visualization: “Jack n Coke”: Lev Golod, Johghyun Lee

Best Use of External Data: “JSS”: Shahid Cohan, James Lepore, Sean Grant 

Read more about DataFest 2014 here:

GridPoint Provides Dataset for National College Tournament “DataFest” 

The Students Most Likely to Take Our Jobs
fivethirtyeight.com

A Weekend of Data Crunching
Duke Magazine

DataFest Sees Teams Competing in Data Analysis
The Daily Princetonian