Grasp the Momentum: Define and Prediction Models for Accurate Trend

Abstract

In tennis, the concept of momentum is used to describe the reversal of a player's big lead. However, motivation in sports is subjective and difficult to quantify. To address this challenge, the study aims to build models that can predict momentum and successful streaks. We accomplished three main tasks. Task one involves assessing player motivation using a variety of metrics and analytic hierarchy process (AHP). Task two includes proving strong autocorrelation and testing fringes using run tests. The results show significant autocorrelation, which is suitable for momentum regression prediction. In Mission three, the random forest method is used to calculate the importance of features to provide suggestions for the player's next game. The long short-term memory (LSTM) model is used to predict the trend of the next game and the momentum of the players, showing good reliability and accuracy. The sensitivity analysis and robustness verification show that it is consistent with the actual situation and the error range is acceptable. Finally, the advantages and disadvantages of the model are summarized, and suggestions for further improvement are put forward.