Datathon Challenges Students to Create Solutions to Real-World Problems


The Institute for Data Engineering and Science (IDEaS) teamed with Correlation One and sponsors Citadel LLC and Citadel Securities to host a Datathon, or Data Open, in the Klaus Advanced Computing Building on Feb. 3.  Faculty members Polo Chau (School of Computational Science and Engineering), Guillermo Goldsztein (School of Mathematics), and Dana Randall (School of Computer Science and co-executive director of IDEaS) served as judges.  Nearly 500 students applied to participate, and the top 100 were invited to compete. At the competition, aspiring data scientists analyzed a problem statement and data sets related to the future of urban environments and "smart" cities. Students competed in teams of four for a $25,000 cash prize.
Four students from Georgia Tech took home the grand prize.  Their analysis reviewed the impact of public service building locations on traffic flow patterns.  Through their analysis, the team determined a lack of hospitals in highly populated areas in Atlanta. They proposed the strategic development of an additional hospital in the southern part of Atlanta, in an area with low traffic incident reports which would address the limited access to hospitals throughout the city.

1st place: Team 6: "Data Informed Urban Planning"
Rohit Mittapalli
Brandon Bakr
David Giles
Benedikt Brandt

2nd place: Team 15: "An Analysis of Government Investing Towards Improving the Transportation Sector"
Xinyi Liu
Rahul Patel
Mayank Kishore
Avinash Ajit Nargund

3rd place: Team 3: "Metropolis: A study of hazards in large cities in the United States"
Nitish Katta
Kshitij Kulkarni
Will Epperson
Angel Alex Cabrera

The Data Open at Georgia Tech is part of a series of 20 Datathons co-hosted by Correlation One and Citadel this year. Datathons are held at the most prestigious technical universities in the world, including Stanford University, Massachusetts Institute of Technology, Oxford University, Peking University, Ecole Polytechnique, the University of Toronto, and many more.

“Datathons help us to identify the world’s best data science talent through a rigorous and scientific process.”  Sham Mustafa, co-founder and CEO of Correlation One said.  “They also provide students who participate a truly meaningful and rewarding learning experience.”

The Data Open simulates the problems that data scientists face on the job.  It assesses a candidate’s ability to pose meaningful questions, work under pressure, work with others and use cutting-edge techniques to extract insight from data. Upcoming competitions are listed on Citadel's Data Open page.

The top three winning teams, faculty judges, and event sponsors at Datathon 2018. From left to right: Kshitij Jain (Correlation One), Petar Todorovic (Correlation One), David Giles (1st place), Guillermo Goldsztein (Judge/sitting), Brandon Bakr (1st place), Will Epperson (3rd place), Vikram Kejariwal (Correlation One/sitting), Sham Mustafa (Correlation One/sitting), Benedikt Brandt (1st place/standing behind), Xinyi Liu (2nd place/standing), Nitish Kulkarni (3rd place/standing behind), Avinash Nargund (2nd place/standing front center), Kshitij Kulkarni (3rd place/standing behind), Rahul Patel (2nd place/ sitting), Mayank Kishore (2nd place/ standing right center), Angel Alex Cabrera (3rd place/ standing behind), Prof. Randall, Prof. Aluru, Prof. Chau.

The first place team: David Giles, Brandon Bakr, Benedikt Brandt, and Rohit Mittapalli

One hundred students were selected from nearly five hundred applicants to compete at Datathon 2018.

Team #15, which placed second in the competition, discussing their approach to analyzing government investments in transportation.


Jennifer Salazar, IDEaS Director of Communication and Grant Writing