A Method Based on LightGBM for Analyzing and Predicting the Outcomes of Tennis Matches
A Method Based on LightGBM for Analyzing and Predicting the Outcomes of Tennis Matches
Wenyu Liu,Xi Xu,Yongqiang Cui,Di Bai
TLDR
A LightGBM-based method is implemented to provide a scientific basis for understanding and predicting the scoring ability of tennis players in matches, which is significant for the field of sports science and technology.
Abstract
The accurate assessment of athletes' tactical performance is crucial for success in highly technical and competitive sports. Comprehensive modeling of tennis players' tactical scoring ability has become a key direction of current academic research due to the continuous advancement of sports technology. This study implements a LightGBM-based method to analyse and predict the situation of tennis matches. Machine learning algorithms such as LGBM, XGBOOST, support vector machines, perceptron machine networks, and logistic regression were used to compare the model results. The LGBM method with the best performance was selected for modelling to achieve dynamic assessment of players' performance. The accuracy, precision, recall, F1 score, and AUC were 0.69, 0.69, 0.7, 0.69, and 0.77, respectively. This research method aims to provide a scientific basis for understanding and predicting the scoring ability of tennis players in matches, which is significant for the field of sports science and technology.