网球比赛结果预测模型的构建  

Construction of Tennis Match Result Prediction Model

作　　者：张沛潇 Zhang Peixiao(College of Electronic Science and Engineering,Jilin University,Jilin,China)

机构地区：[1]吉林大学电子科学与工程学院,吉林长春

出　　处：《科学技术创新》2024年第19期33-36,共4页Scientific and Technological Innovation

摘　　要：本文依据2024年美国大学生数学建模竞赛C题,并借鉴2023年温布尔登男子网球公开赛决赛结果数据,进行了比赛优异模型及波动模型的构建及应用,旨在为运动员和教练提供更为精准的训练及比赛策略。首先通过建立比赛优异模型对势头进行量化,并验证该模型的鲁棒性;其次建立比赛波动模型模拟球员的心理状态,模型准确率为72.9%,并分析准确率未达到理想状态的原因,提出相应的建议;最后对模型进行泛化,探讨将其应用于其他比赛的可行性。Based on Question C of the 2024 American College Students Mathematical Contest in Modeling and the data of the 2023 Wimbledon Men's tennis final,this paper constructs and applies the competition excellence model and fluctuation model,aiming to provide more accurate training and competition strategies for athletes and coaches.Firstly,the momentum is quantified by establishing a competition excellence model,and the robustness of the model is verified.Secondly,a game fluctuation model is established to simulate the players'mental state,and the accuracy rate of the model is 72.9%,and the reasons why the accuracy rate does not reach the ideal state are analyzed,and the corresponding suggestions are put forward.Finally,the model is generalized and the feasibility of applying it to other competitions is discussed.

关 键 词：网球 SelectKBest模型 游程检验 Adaboost模型 GBDT模型 

分 类 号：G845[文化科学—体育训练]