Loyola mini-Datafest Competition
- April 10, 2015
- 12:30 PM - 4:00 CST
- , Room 123
Computer Science and Mathematics & Statistics
Earvin Balderama, firstname.lastname@example.org
All current Loyola students and staff
Open to the public.
- Add to calendar
This semester students are encouraged to compete in our Loyola data analytics competition (a.k.a. LUC mini-Datafest) cosponsored between Computer Science and the Department of Mathematics and Statistics. All students and staff at Loyola are free to compete individually or in pairs. The top 3 groups will be awarded from $50 up to $150. You are encouraged to register here.
This Spring's competition is a prediction/machine learning challenge in the style of a kaggle.com competition. The chosen data set won't be as fun as dogs vs. cats or as lucrative as the netflix prize, but we think you'll enjoy it.
To help enrich the field, the event will begin with a one hour tutorial. Everyone will be shown how to obtain a baseline result on their laptops with provided sample data and be given pointers on ways to improve their results.
After that, it's a free-for-all to see who can improve their predictions and get the highest accuracy.
Note: we call this *mini-* in comparison to longer full-day competitions like the Kraft data dive taking place in Chicago April 18th (btw, registration for that event is due THIS FRIDAY, March 20th, for anyone interested)
LUC Datafest Details:
Date: Friday, April 10th
Time: 12:30 - 4pm (1 hour tutorial followed by 2 hour timed competition)
Location: Institute of Environmental Sustainability (IES) 123
Participation: individually or in pairs, registration encouraged at http://datafest.pacsites.org, but walk-ins are welcome too if there is space.
Laptops: Students are expected to bring laptops with wireless internet connectivity
Setup: the tutorial will use the free Anaconda python distribution, but any programming language can be used for analysis (no GUI-only software). Data will be made available in csv format. Results will be submitted online and scored automatically.
Judging: A leaderboard will be updated during the competition giving accuracy on a released validation set. Final results may be subject to additional validation. Highest score wins.