Guidelines

What is the Datathon?

0
votes

The Datathon is a weekend-long online competition where you are challenged to work on a real-world business case from different areas of Machine Learning, AI, and Data Science.

The teams have 48 hours to come up with a solution to a real business challenge based on the provided datasets. The jury and the society will award the most precise, but also creative solutions.

The Datathon is entirely held online and opened to the global community via our platform. Anyone who wants to join us in any of our Data Hubs is welcome.

The Datathoners are part of the Data Science Society – a community for all Data lovers, experts and geeks to learn from each other and to apply their knowledge in international competition at the same weekend. The Datathon is happening at the same time [different time zones possible] and shares the same data cases as challenges, the results of which are published over the central Data platform of Data Science Society.

During the Datathon there are 4 distinctive roles – participants (who form teams and solve the cases), mentors (experts who help the teams in solving the cases), industry experts (representatives from the companies which provide the cases with more domain knowledge) and hosts (organizations who locally organize Datathon Hubs).

The detailed case descriptions of the Datathon is announced for the participants to choose their challenge at least 1 week before the Datathon itself, the data sets of the chosen challenge are published on the first day (Friday) of the Datathon. The results must be shared by announced in the schedule time (local time) after two days (on Sunday).

Teams are formed virtually up until the start of the Datathon. For participants without teammates the Matchmaker suggests teams. Each team then would choose which case to work on and start working.

The DSS platform and especially the DataChat are the main place of communications among participants, teams, organizers, mentors and industry experts during the Datathon.

The teams can use the output data and other information provided to them in regard to solving the chosen case only during the duration of the event. They must not use those for personal or commercial purposes outside the event format and activities. With the registration for the Datathon, the teams agree to also provide their work (analysis, methodology, source code, etc) to the organizers for them to publish on the event page.

The teams submit their work as an article when the 48 hours of the challenge are over. Every team’s solutions will be evaluated according to several criteria – statistical indicators, resourcefulness of the solution and creativity of visualization.

The event finishes with the mentors jurying all the team solutions by voting. For the Semi-finals the teams will be assigned in groups of 4-5 (working on the same case, and maybe one group with various cases) on Saturday evening. Every group gets to be judged by one mentor, one domain expert (from the company which provides the case) and one DSS expert. The jury ranks all teams within the group in ascending order of how good was their solution (rank 1 means “the best”). The team which gets ranked at the top of their group qualifies for the final.

The finalist teams would have a live 3 minute presentation (over Google hangout) followed by 2 minutes Q&A session by the Jury in the General Data.Chat channel. For this round the Jury would consist of only Mentors and DSS experts all of whom get to rank the finalists (again in a provided survey form). The team with the best collective rank wins.

Share this

Leave a Reply