Machine Learning-Based Prediction of Olympic Medals and The Exploration of National Sports Expertise

This study focuses on research related to the Olympic Games, employing a combination of Analytical Hierarchy Process (AHP), Random Forest Regression, and K-means Clustering techniques to analyze athletes' abilities, medal outcomes of various countries, and their sporting specializations. First, an AHP-based athlete performance evaluation model is constructed, where athletes are scored based on total medal count, gold medal count, number of events participated in, and event diversity. Second, a Random Forest Regression prediction model is developed, which, using historical medal data of countries and athletes' scores, forecasts the medal outcomes for different countries in the 2028 Olympic Games. The model demonstrates strong predictive accuracy. Finally, K-means Clustering is applied to categorize countries, exploring the sports in which different clusters of countries excel. The findings of this research provide scientific evidence for formulating national sports development strategies and analyzing the relationship between sports and various factors, thus contributing to enhancing national sports competitiveness and advancing the development of sports.