Bayesian-LightGBM Classification Prediction Model of Tennis Match Situation Change

In order to determine when the flow of a tennis match is about to change in favor of one player to another, a Bayesian-LightGBM classification prediction model for tennis match situation change is proposed in this paper. Seven dynamic parameters of one player in a tennis match are taken as the characteristic parameters of the change of the competition situation, it is classified and quantified to generate the discrete data set to meet the needs of the model. Driven by this data set, a Bayesian-LightGBM algorithm seven classification model of tennis match situation changes is established, which can effectively identify seven kinds of match situation changes (highly favorable, relatively favorable, favorable, tie, unfavorable, relatively unfavorable, highly unfavorable), with an accuracy of more than 90%. The results show that the LightGBM classification model can effectively identify the situation changes of 7 kinds of tennis matches and has superior robustness. In addition, through the analysis of the importance of parameters, it is found that the number of consecutive wins and the first-serve scoring rate of players have a great influence on the change of tennis match situation.