Models for Predicting Olympic Medal Tables

Abstract

During the 2024 Paris Olympics, the medal table attracted intense attention, and there was a surge of enthusiasm for predicting the medals of the 2028 Olympics. This study focuses on constructing a model to indicate the number of Olympic medals, aiming to explore the factors influencing a country's performance in winning Olympic medals and make accurate predictions. First, we conducted rigorous data cleaning operations on the collected data to ensure that the data input into the model met the requirements. Then we adopted a two-stage model construction approach to predict the number of medals these countries could obtain. The virtual medal prediction leaderboard shows that Bangladesh and Benin might win their first Olympic medals in 2028. Next, we used the Kmeans++ clustering model to group similar countries together. We also analyzed the relationship between sports events and medal counts. We identified their advantageous sports events and visualized the clustering results through data visualization. Overall, this model provides a practical tool for reasonably predicting the number of medals. By considering multiple factors, it offers insights for countries on how to win more medals.