基于组合模型的球员贡献度评价  

Player contribution evaluation based on combination model

作　　者：代浩然 曹文芹 DAI Haoran;CAO Wenqin(School of Mathematics and Statistics,Shandong University of Technology,Zibo 255049,China)

机构地区：[1]山东理工大学数学与统计学院,山东淄博255049

出　　处：《山东理工大学学报（自然科学版）》2024年第3期71-78,共8页Journal of Shandong University of Technology:Natural Science Edition

摘　　要：介绍了两种评价球员贡献度的统计模型,基于组合模型的优势,提出了1个组合评价模型,以2021—2022年NBA季后赛为例,利用非参数统计中Friedman检验进行了实证分析。通过基于熵权法改进的Topsis、主成分综合评价与组合模型3种评价模型来探究2021—2022年NBA总冠军金州勇士队球员的贡献,并对其贡献进行排名,采用Friedman检验得到3种模型的结论基本一致。In the paper,we firstly introduce two statistical models for evaluating player contribution,and then propose a new combined evaluation model based on the advantages of the combination model.We finally performed an empirical analysis using nonparametric statistical Friedman test for the 2021—2022 NBA playoffs.Golden state Warriors is the champion team,and the team is analyzed as an example using improved entropy weighting method,principal component analysis,and the combined model.We explore the contributions of Golden State Warriors players and rank their contributions.We come to a conclusion that the three models are consistent according to the Friedman test.

关 键 词：非参数统计 相关分析 组合模型 Friedman检验 

分 类 号：O213[理学—概率论与数理统计]