The Olympic Datathon is designed to be a language-agnostic mini "hackathon" for students with varying levels of experience. The focus of the datathon will be the exploratory data analysis - or as we like to call it data storytelling - of the given Olympics dataset*. We invite everyone who is interested in data science to participate, regardless of their technical background. This one day datathon offers an interesting dataset to explore, guidance along the way, and free pizza!
Disclaimer: This is an EDA and not a ML competition.
*Dataset will be published on October 16th.
Spotted a mistake, dead link, or have suggestions for improvements? Message us using the links below.
- Make sure you have registered on the Google form. Otherwise, you will not be invited/accepted to the Microsoft Teams group.
- Join the Datathon Teams Group and turn on notifications to stay up to date. The hybrid session will be hosted there.
- Click 'Going' on the datathon's Facebook event and invite your friends if you want.
All other relevant details should be available on this page. If you have any questions feel free to reach out by using the contact links below. Viel Spass!
All events on October 16th will be held both in-person as well as online. For in-person attendees, please join us in room OC0.01. Online attendees can follow the in-person event on the 'General' channel on Teams. We strongly encourage anyone who has the chance to participate in the in-person event.
Again, the dataset will be published on October 16th.
| Date | Time | Event |
|---|---|---|
| 16th October | 3-4pm | Dataset introduction & Python basics for storytelling |
| 16th October | 4-5pm | R workshop by Warwick Statistics Society |
| 16th October | 5-6pm | In person guidance |
| 16th October | 6pm | Free pizza |
| 23rd October | 6pm | Submission deadline |
Which factors determine whether an Olympic athlete wins a medal or not?
Congratulations to the team consisting of Ammar Salem, Dulnath Jayasinghe, Sofia Pangher and Zain Mobarik!
- Summary statistics
- Univariate analysis (e.g. Medal)
- Bivariate analysis (e.g. Height vs Weight)
- Statistical modelling and inference (If you feel brave enough. This is not part of the EDA!)
Make sure to elaborate on your observations!
- Google colab skeleton code. Save a copy in Drive and follow along in the teaching session.
- Data set
- Python for Data Science teaching session 1 by WDSS
- Tidyverse R basics
Submissions should be sent to hackathon@wdss.io. We accept any formats!
The competition will run from 6pm on Saturday the 16th October to 6pm on Saturday the 23rd October. No preference will be given to early entrants.
We will judge submissions based on the insights they derive. If you can tell a story about the data with basic code, but elaborate observations that's actually slightly better than a super technically advanced submission with no elaborations at all. We will be looking for a good balance between code, visualizations and elaborations.
The winner/winning team will receive a shoutout on our social media accounts (including LinkedIn if you wish so). We will also link to your solution from this GitHub page.
This is up to you to decide. If you would like to be part of a team, please contact Valentin Kodderitzsch on MS teams. We will try to our best to allocate teams of up to 4 people based on each participant's skill level. We will also be discussing teams during the Datathon introduction workshop which will be held on October 16th 2021 at 3pm in OC0.01.
No problem. Simply email hackathon@wdss.io and explain the situation.
We will provide help throughout the datathon in a number of ways:
- Most help will be provided during the workshop on October 16th
- After the workshop you will find help by messaging the 'Get Help' channel on Teams
This hackathon was written and is primarily managed by Valentin Kodderitzsch. If you need to contact me please reach out on Microsoft Teams or message me on LinkedIn.