THE PREDICTION OF OLYMPIC MEDAL TABLE BASED ON LINEAR REGRESSION MODELING

As the world’s most influential sporting event bringing together the world's best athletes, the Olympic Games is the highest stage for competitive sports. It inspires more people to participate. In this paper, in order to predict the total medals and gold medals won by each country in the 2028 Olympic Games, a multiple linear regression model is constructed by considering the historical medal datas, the number of athletes' participation and the types and number of participating events and other characteristic variables as the indexes, and takes the evaluation coefficient R2and the mean squared error MSE as the model evaluation indexes. Through the established medal list and historical trend, the prediction interval of total medals and the prediction interval of gold medals are analyzed, and those countries that may progress or regress in the 2028 Olympic Games are analyzed and obtained, and by the prediction this paper selects the top 10 progressing countries and the 10 countries with obvious regression. In addition, the prediction of those countries that have never won a prize is also made to explore the possibility of winning a medal, and for this purpose, this paper adopts a binary classification model and logistical regression model, and the probabilities of winning a first medal are obtained by selecting the data of the countries that have never won the award.