A Study of Momentum Prediction in Tennis Based on Optimization Model, Prediction Model And LSTM

Abstract

In the highly anticipated 2023 Wimbledon men's singles final, 20-year-old Spanish star balllos Alballaz faced off against the experienced 36-year-old champion Novak Djokovic. The match had its ups and downs, with Djokovic dominating the first set in a big way, however, things changed at the start of the second set, and the match eventually went all the way to the fifth set, where Alballas was victorious 6-4. The provided dataset, named "2023-wimbledon-1701", contains detailed data on every shot in the game, providing a valuable resource for understanding game dynamics and momentum shifts. In this analysis, paper aim to explore the factors that change momentum during a match, taking into account serve probability, and incorporate it into the model. Integrate different explicit indicators according to the new evaluation model, simulate the regression curve for key explicit indicators, observe the deviation value, and according to the deviation value, establish a new model to predict the selected data, and use MATLAB to continuously simulate and train to ensure that the data is within the prediction interval. Our goal is to gain a comprehensive understanding of how events within a game influence the creation or change of momentum. Ultimately, the model-based visualization will provide us with deep insights into the detailed and dynamic nature of the Wimbledon men's singles final. Predicting the effect of momentum changes on the outcome of a race by capturing the details and detecting the nature of the dynamics. Overall, the model shows that the relationship of new joint distribution sports and covariance function on sports cross.