Research on Medal Prediction Model for 2028 Olympic Games Based on Linear Regression and Random Forests

Abstract

This paper proposes an Olympic medal prediction model integrating linear regression, random forest, LASSO-Logistic regression and hierarchical analysis method (AHP), focusing on the application of multiple models in the prediction of the number of medals, the analysis of the probability of winning awards and the quantification of the effect of the “great coach”. Firstly, based on linear regression, the index system is constructed, combined with the least squares method and the random forest algorithm, the number of gold, silver and bronze medals of each country in the 2028 Olympic Games is predicted at the point prediction and the interval prediction; secondly, the LASSO-Logistic regression model is introduced, and the probability of winning each item is analyzed by screening the characteristics of the penalty parameter; lastly, the “Great Coach” effect analysis model is established by using the AHP. Finally, AHP was used to establish the “Great Coach” effect analysis model to quantify the weight of its influence on athletes' training intensity, team tactics and other factors. The model synthesizes historical medal data, participation scale and other multi-dimensional indicators, and improves prediction accuracy through algorithmic fusion, which can scientifically assess the medal distribution trend and key influencing factors, providing a multi-method synergistic solution for Olympic medal prediction.