Carolina Data Challenge

Congratulations to the team of Supreet Deshpande, Chris Goodrich, Taylor Kooy, and Meghan Weber, on taking home the 1st place prize for “Best Insight” at the 2019 Carolina Data Challenge, an annual event organized by students in the Analytics and Data Science Club at the University of North Carolina at Chapel Hill. The challenge brought together 250 competitors, organized into 51 teams, working over a 24-hour period. The Institute’s team, competing in the insight category, performed a time series analysis to determine how the recent agricultural trade war has affected U.S. farmers exporting sorghum. 

Commenting on their win, Supreet observed the key was synergy: “What made the seemingly impossible task achievable was our teamwork, passion, and commitment towards the common goal. It’s astonishing to see what can be achieved in 24 hours—with the right attitude and an extraordinary team.”

Congratulations also go to the team of Jess Conner, Catherine McSorley, Jonathan Rice, and Ryan Weisner, for their contribution in the visualization category, which examined the amount of corn the United States must  produce to meet the growing consumer demand for beef. The team’s R shiny dashboard uses vectorized ARIMA to forecast corn production. It also visualizes the amount of land needed for total corn production, and corn production specifically for feeding cattle, projected geographically across the mainland U.S.

“I had a lot of fun being able to work with my talented team in this competition!,” said Catherine. “Implementing and expanding on tools we learned in our coursework was really rewarding for all of us, and while the 24 hour time pressure was a challenge, it also sparked our creativity.”

The Institute has been a regular sponsor of the Carolina Data Challenge since its inception in 2017.

 

Insights team (L-R): Taylor Kooy, Supreet Deshpande, Meghan Weber, and Chris Goodrich.
Visualization team (L-R): Jess Conner, Ryan Weisner, Jonathan Rice, and Catherine McSorley.
Data visualization dashboard by Conner, Weisner, Rice, and McSorley.