Olympic Sages by DataDruids

We are Team Data Druids, united across drastically different time zones to analyze data and predict the winners of the 2024 Paris Olympics. A big part of our analysis was done on a dataset from the 2021 Tokyo Olympics. We began by hypothesizing how past Olympic data could offer valuable insights, focusing on which countries won the most medals.

We then conducted a deep dive into the dataset, cross-referencing various categories to draw valid conclusions. This included identifying the total number of athletes and coaches each country had, as well as considering other factors such as the genders of the athletes and the distribution of medals won by each country.

After visualizing our results, we discussed potential predictions. Initially, we used a linear regression model to predict the countries likely to win the most medals in the 2024 Olympics. However, this model did not perform well, as indicated by the mean absolute error and R-squared values.

We then switched to a random forest regressor model, incorporating more data, and achieved much better results. Our insights and visualizations were used to build a webpage to showcase all our findings. We are committed to continuing this project beyond the hackathon to improve predictions for future Olympics.