Olympic Medal Count Prediction Model for Various Countries based on LSTM and Supervised Machine Learning

The acquisition of Olympic medals holds significant importance for the development of a country's sports endeavors. This paper constructs a medal prediction model based on TOPSIS-LSTM model and supervised learning, utilizing historical Olympic data. The Random Forest algorithm is employed to forecast the medal performance of countries at the 2028 Los Angeles Olympics. The results indicate that the United States will achieve 126 medals, while China will secure 91 medals, ranking first and second, respectively. The United Kingdom and Canada follow closely with 65 and 55 medals, respectively. The model's RMSE is less than 5.8, and the R2 value is greater than 0.93, indicating a relatively good fit.