Prediction and analysis of Olympic medal distribution based on random forest multifactor model

This paper aims to build a random forest multi-factor model to investigate the distribution of Olympic MEDALS in depth. Using the data set of historical Olympic MEDALS, the research integrates the external characteristics and lagging characteristics of the country such as population and GDP through data processing and feature engineering. The random forest regressor was used to predict the 2028 Olympic medal distribution, the hyperparameters were optimized by GridSearchCV, and the model performance was evaluated by mean square error (MSE) and determination coefficient. At the same time, the random forest classifier was used to predict the probability of winning the first medal of the non-winning country. The study found that the historical medal data is the key factor to predict the number of MEDALS, excellent coaches and core events have a significant impact on the number of MEDALS, and the host effect is mainly reflected in new events and traditional advantageous events. This model provides theoretical and practical support for countries to formulate sports development strategies, help improve Olympic performance and promote sustainable sports development.