Research on Olympic Medal Counts Based on the RF-CUSUM-DID Model

With the advent of the intelligent era, machine learning technology has been widely and deeply applied in the sports industry today, making the development of accurate and comprehensive predictive models crucial. However, existing research on sports event prediction models rarely takes into account both subjective and objective factors comprehensively. To address this challenge, this study uses the random forest algorithm to analyze and predict historical Olympic medal data, achieving an R2 of 53.06% and a MAE of 0.379, outperforming other common machine learning models in comprehensive comparisons. Additionally, a double difference algorithm was incorporated to quantify the impact of renowned coaches on the enhancement of medals, and the CUSUM algorithm was used to test and verify the effect of renowned coaches in historical Olympic events of various countries. The study demonstrates the robustness and effectiveness of the random forest algorithm in handling complex sports prediction tasks, especially in integrating both subjective and objective factors.